Model 8-2x Advanced CNN with DA(cv5).ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np \n",
    "import pandas as pd \n",
    "from subprocess import check_output\n",
    "#print(check_output([\"ls\", \"../input/\"]))\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.metrics import log_loss\n",
    "from os.path import join as opj\n",
    "from matplotlib import pyplot as plt\n",
    "import tensorflow as tf\n",
    "import os\n",
    "%matplotlib inline\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from matplotlib import pyplot\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Conv2D, MaxPooling2D, Dense, Dropout, Input, Flatten, Activation,Add, ZeroPadding2D ,AveragePooling2D\n",
    "from keras.layers import GlobalMaxPooling2D\n",
    "from keras.layers.normalization import BatchNormalization\n",
    "from keras.layers.merge import Concatenate\n",
    "from keras.models import Model\n",
    "from keras import initializers,regularizers\n",
    "from keras.optimizers import Adam,SGD\n",
    "from keras.initializers import glorot_uniform\n",
    "from keras.callbacks import ModelCheckpoint, Callback, EarlyStopping\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "import keras.backend as K\n",
    "K.set_image_data_format('channels_last')\n",
    "#K.set_learning_phase(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = pd.read_json(\"input/train.json\")\n",
    "#test = pd.read_json(\"input/test.json\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "only_band1 = ['band1_'+str(i) for i in range(75*75)]\n",
    "only_band2 = ['band2_'+str(i) for i in range(75*75)]\n",
    "both_bands = only_band1+only_band2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df = pd.concat([pd.DataFrame(test['band_1'].values.tolist(),columns = only_band1),\n",
    "#                 pd.DataFrame(test['band_2'].values.tolist(),columns = only_band2)],axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def collect_test():\n",
    "    test_data = pd.DataFrame()\n",
    "    for i in range(4):\n",
    "        test_piece = pd.read_csv('input/test_subparts/test'+str(i+1)+'.csv')\n",
    "        test_data = pd.concat([test_data,test_piece])\n",
    "        test_data = test_data.reset_index().drop(['index'],axis = 1)\n",
    "    return test_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_data = collect_test()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_band_test_1=np.array(test_data[only_band1]).reshape(8424,75,75)\n",
    "X_band_test_2=np.array(test_data[only_band2]).reshape(8424,75,75)\n",
    "X_test = np.concatenate([X_band_test_1[:, :, :, np.newaxis], X_band_test_2[:, :, :, np.newaxis], ((X_band_test_1+X_band_test_2)/2)[:, :, :, np.newaxis]], axis=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_band_1=np.array([np.array(band).astype(np.float32).reshape(75, 75) for band in train[\"band_1\"]])\n",
    "X_band_2=np.array([np.array(band).astype(np.float32).reshape(75, 75) for band in train[\"band_2\"]])\n",
    "X_train = np.concatenate([X_band_1[:, :, :, np.newaxis], X_band_2[:, :, :, np.newaxis],((X_band_1+X_band_2)/2)[:, :, :, np.newaxis]], axis=-1)\n",
    "target_train=train['is_iceberg']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# X_band_test_1=np.array([np.array(band).astype(np.float32).reshape(75, 75) for band in test[\"band_1\"]])\n",
    "# X_band_test_2=np.array([np.array(band).astype(np.float32).reshape(75, 75) for band in test[\"band_2\"]])\n",
    "# X_test = np.concatenate([X_band_test_1[:, :, :, np.newaxis], X_band_test_2[:, :, :, np.newaxis], ((X_band_test_1+X_band_test_2)/2)[:, :, :, np.newaxis]], axis=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "X_train_cv, X_valid, y_train_cv, y_valid = train_test_split(X_train, target_train, random_state=2,stratify = target_train, train_size=0.80)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Cross Validation Folds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def data_augmentation(X_input,Y_input,batch_size = 32):\n",
    "    data_aug = ImageDataGenerator(#featurewise_center=True,\n",
    "                             #featurewise_std_normalization=True,\n",
    "                             #zca_whitening=True,\n",
    "                             rotation_range=0,\n",
    "                             width_shift_range = 0,\n",
    "                             height_shift_range = 0,\n",
    "                             zoom_range = 0,\n",
    "                             data_format = 'channels_last',\n",
    "                             horizontal_flip = True,\n",
    "                             vertical_flip = True)#,fill_mode = 'constant',cval = 0)\n",
    "    data_aug_batches = data_aug.flow(X_input,Y_input,batch_size = batch_size)\n",
    "    return data_aug_batches\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def cnnmodel():\n",
    "    X_input = Input(shape = (75,75,3))\n",
    "    #First composite layer\n",
    "    X = Conv2D(128,kernel_size = (5,5))(X_input)\n",
    "    X = Activation('relu')(X)\n",
    "    X = MaxPooling2D(pool_size=(3,3),strides =(2,2))(X)\n",
    "    X = Dropout(rate = 0.2)(X)\n",
    "    \n",
    "    #Second composite layer\n",
    "    X = Conv2D(256,kernel_size = (3,3))(X)\n",
    "    X = Activation('relu')(X)\n",
    "    X = MaxPooling2D(pool_size=(3,3),strides =(2,2))(X)\n",
    "    X = Dropout(rate = 0.2)(X)\n",
    "    \n",
    "    #Third composite layer\n",
    "    X = Conv2D(256,kernel_size = (3,3))(X)\n",
    "    X = Activation('relu')(X)\n",
    "    X = MaxPooling2D(pool_size=(2,2),strides =(2,2))(X)\n",
    "    X = Dropout(rate = 0.2)(X)\n",
    "\n",
    "    #Forth composite layer\n",
    "    X = Conv2D(128,kernel_size = (3,3))(X)\n",
    "    X = Activation('relu')(X)\n",
    "    X = MaxPooling2D(pool_size=(3,3),strides =(2,2))(X)\n",
    "    X = Dropout(rate = 0.2)(X)\n",
    "\n",
    "    #Flatten layer\n",
    "    X = Flatten()(X)\n",
    "    \n",
    "    #First dense layer\n",
    "    X = Dense(512,activation = 'relu')(X)\n",
    "    X = Dropout(rate = 0.2)(X)\n",
    "   \n",
    "    #Second dense layer\n",
    "    X = Dense(256,activation = 'relu')(X)\n",
    "    X = Dropout(rate = 0.2)(X)\n",
    "   \n",
    "    #Decision layer\n",
    "    X = Dense(1,activation='sigmoid')(X)\n",
    "    \n",
    "    model = Model(inputs=X_input,outputs=X)\n",
    "\n",
    "    return model\n",
    "\n",
    "def model_compile(model,lr = 0.0001,decay = 0, freezing_layers = None,weights_path = None):\n",
    "   \n",
    "    if weights_path:\n",
    "        model.load_weights(weights_path)\n",
    "    else:\n",
    "        pass\n",
    "    if freezing_layers:\n",
    "        for layer in model.layers[:freezing_layers]:\n",
    "            layer.trainable = False\n",
    "    else:\n",
    "        pass\n",
    "    optimizer = Adam(lr = lr, beta_1=0.9, beta_2=0.999, epsilon=1e-08, decay=decay)\n",
    "    #optimizer = SGD(lr, decay=0, momentum=0.9, nesterov=True)\n",
    "    model.compile(loss='binary_crossentropy',optimizer=optimizer,metrics=['accuracy'])\n",
    "    model.summary()\n",
    "    return model\n",
    "\n",
    "def get_callbacks(filepath = \".model_weights.hdf5\", patience=7, monitor = 'val_loss', verbose = 1, save_best_only = True):\n",
    "    es = EarlyStopping('val_loss', patience=patience, mode=\"min\")\n",
    "    msave = ModelCheckpoint(filepath, monitor=monitor, verbose = verbose, save_best_only=True)\n",
    "    return [es,msave]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def normal_fit(model,callbacks,epochs = 100,batch_size = 32,verbose = 1):\n",
    "    \n",
    "    result = model.fit(X_train_cv, y_train_cv,\n",
    "                          batch_size=batch_size,\n",
    "                          epochs=epochs,\n",
    "                          verbose=verbose,\n",
    "                          validation_data=(X_valid, y_valid),\n",
    "                          callbacks=callbacks)\n",
    "    return result\n",
    "\n",
    "def data_augment_fit(model,callbacks,epochs = 100,batch_size = 32,verbose = 1):\n",
    "    \n",
    "    result = model.fit_generator(data_augmentation(X_train_cv,y_train_cv,batch_size = batch_size),\n",
    "                          steps_per_epoch = len(X_train_cv)/batch_size,\n",
    "                          epochs=epochs,\n",
    "                          verbose=verbose,\n",
    "                          validation_data=(X_valid, y_valid),\n",
    "                          callbacks=callbacks)\n",
    "    return result\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Cnnmodel = model_compile(cnnmodel(),lr = 0.0001,decay = 0, freezing_layers = None,weights_path = None)\n",
    "callbacks = get_callbacks(filepath = 'model save/Model 7-2x filters Advanced CNN with DA and self TL/2018.1.2.hdf5', patience=100, save_best_only = False)\n",
    "result = normal_fit(Cnnmodel,callbacks = callbacks,epochs = 100,batch_size=32,verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Cnnmodel = model_compile(cnnmodel(),lr = 0.0001,decay = 0, freezing_layers = 5,weights_path = 'model save/Model 7-2x filters Advanced CNN with DA and self TL/2018.1.2.hdf5')\n",
    "callbacks = get_callbacks(filepath = 'model save/Model 7-2x filters Advanced CNN with DA and self TL/2018.1.2_stage2.hdf5', patience=30, save_best_only = True)\n",
    "result = data_augment_fit(Cnnmodel,callbacks = callbacks,epochs = 100,batch_size=32,verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "Cnnmodel.load_weights('model save/Model 7-2x filters Advanced CNN with DA and self TL/Model-7 cv1_stage1.hdf5')\n",
    "Cnnmodel.evaluate(X_valid, y_valid)\n",
    "#Cnnmodel.evaluate(X_train_cv, y_train_cv)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def save_history(result,name = 'fit_history.csv',path = os.getcwd()):\n",
    "    train_loss = result.history['loss']\n",
    "    val_loss = result.history['val_loss']\n",
    "    train_accuracy = result.history['acc']\n",
    "    val_accuracy = result.history['val_acc']\n",
    "    fit_history = pd.DataFrame({'train_loss':train_loss,'val_loss':val_loss,'train_accuracy':train_accuracy,'val_accuracy':val_accuracy})\n",
    "    fit_history.to_csv(os.path.join(path,name))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "save_history(result_Cnn)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Model7_CV(X_train,K=3):\n",
    "    folds = list(StratifiedKFold(n_splits=K, shuffle=True,random_state = 3).split(X_train, target_train))\n",
    "    y_train_pred_log=0\n",
    "    y_valid_pred_log = 0.0*target_train\n",
    "    for j, (train_idx, valid_idx) in enumerate(folds):\n",
    "        print('\\n===================FOLD=',j+1)\n",
    "        global X_train_cv\n",
    "        X_train_cv = X_train[train_idx]\n",
    "        global y_train_cv\n",
    "        y_train_cv = target_train[train_idx]\n",
    "        global X_valid\n",
    "        X_valid = X_train[valid_idx]\n",
    "        global y_valid\n",
    "        y_valid= target_train[valid_idx]\n",
    "        \n",
    "        #Angle\n",
    "#         X_angle_cv=X_angle[train_idx]\n",
    "#         X_angle_hold=X_angle[test_idx]\n",
    "\n",
    "        #define file path and get callbacks\n",
    "        \n",
    "        file_path_stage2 = 'model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv'+str(j+1)+'.hdf5'\n",
    "        Cnnmodel = model_compile(cnnmodel(),lr = 0.0001,decay = 0, freezing_layers = None,weights_path = None)\n",
    "        callbacks = get_callbacks(filepath = file_path_stage2, patience=30, save_best_only = True)\n",
    "        result2 = data_augment_fit(Cnnmodel,callbacks = callbacks,epochs = 150,batch_size=32,verbose=2)\n",
    "        print('Satge 2,done!')\n",
    "#         callbacks = get_callbacks(filepath=file_path, patience=5)\n",
    "#         gen_flow = gen_flow_for_two_inputs(X_train_cv, X_angle_cv, y_train_cv)\n",
    "#         galaxyModel= getVggAngleModel()\n",
    "#         galaxyModel.fit_generator(\n",
    "#                 gen_flow,\n",
    "#                 steps_per_epoch=24,\n",
    "#                 epochs=100,\n",
    "#                 shuffle=True,\n",
    "#                 verbose=1,\n",
    "#                 validation_data=([X_holdout,X_angle_hold], Y_holdout),\n",
    "#                 callbacks=callbacks)\n",
    "\n",
    "        #Getting the Best Model\n",
    "        Cnnmodel.load_weights(filepath=file_path_stage2)\n",
    "        #Getting Training Score\n",
    "        score = Cnnmodel.evaluate(X_train_cv, y_train_cv)\n",
    "        print('Train loss:', score[0])\n",
    "        print('Train accuracy:', score[1])\n",
    "        #Getting Test Score\n",
    "        score = Cnnmodel.evaluate(X_valid, y_valid)\n",
    "        print('Valid loss:', score[0])\n",
    "        print('Valid accuracy:', score[1])\n",
    "\n",
    "        #Getting validation Score.\n",
    "        pred_valid=Cnnmodel.predict(X_valid)\n",
    "        y_valid_pred_log[valid_idx] = pred_valid.reshape(pred_valid.shape[0])\n",
    "\n",
    "        #Getting Test Scores\n",
    "#         temp_test=Cnnmodel.predict(X_test)\n",
    "#         y_test_pred_log+=temp_test.reshape(temp_test.shape[0])\n",
    "\n",
    "        #Getting Train Scores\n",
    "        temp_train=Cnnmodel.predict(X_train)\n",
    "        y_train_pred_log+=temp_train.reshape(temp_train.shape[0])\n",
    "\n",
    "#     y_valid_pred_log=y_valid_pred_log/K\n",
    "    y_train_pred_log=y_train_pred_log/K\n",
    "\n",
    "    print('\\n Train Log Loss Validation= ',log_loss(target_train, y_train_pred_log))\n",
    "    print(' Test Log Loss Validation= ',log_loss(target_train, y_valid_pred_log))\n",
    "    return y_valid_pred_log"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "===================FOLD= 1\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_1 (InputLayer)         (None, 75, 75, 3)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 71, 71, 128)       9728      \n",
      "_________________________________________________________________\n",
      "activation_1 (Activation)    (None, 71, 71, 128)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 33, 33, 256)       295168    \n",
      "_________________________________________________________________\n",
      "activation_2 (Activation)    (None, 33, 33, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 14, 14, 256)       590080    \n",
      "_________________________________________________________________\n",
      "activation_3 (Activation)    (None, 14, 14, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_4 (Conv2D)            (None, 5, 5, 128)         295040    \n",
      "_________________________________________________________________\n",
      "activation_4 (Activation)    (None, 5, 5, 128)         0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_4 (MaxPooling2 (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dropout_5 (Dropout)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 256)               131328    \n",
      "_________________________________________________________________\n",
      "dropout_6 (Dropout)          (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 1,584,257\n",
      "Trainable params: 1,584,257\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/150\n",
      "Epoch 00001: val_loss improved from inf to 0.69360, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 21s - loss: 1.2316 - acc: 0.5206 - val_loss: 0.6936 - val_acc: 0.4969\n",
      "Epoch 2/150\n",
      "Epoch 00002: val_loss improved from 0.69360 to 0.67323, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.7296 - acc: 0.5374 - val_loss: 0.6732 - val_acc: 0.5528\n",
      "Epoch 3/150\n",
      "Epoch 00003: val_loss improved from 0.67323 to 0.59677, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.6657 - acc: 0.5694 - val_loss: 0.5968 - val_acc: 0.6677\n",
      "Epoch 4/150\n",
      "Epoch 00004: val_loss improved from 0.59677 to 0.56383, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.5953 - acc: 0.6446 - val_loss: 0.5638 - val_acc: 0.6770\n",
      "Epoch 5/150\n",
      "Epoch 00005: val_loss improved from 0.56383 to 0.54980, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.5687 - acc: 0.6518 - val_loss: 0.5498 - val_acc: 0.6801\n",
      "Epoch 6/150\n",
      "Epoch 00006: val_loss improved from 0.54980 to 0.54195, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.5571 - acc: 0.6713 - val_loss: 0.5419 - val_acc: 0.7019\n",
      "Epoch 7/150\n",
      "Epoch 00007: val_loss improved from 0.54195 to 0.52693, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.5484 - acc: 0.6708 - val_loss: 0.5269 - val_acc: 0.7547\n",
      "Epoch 8/150\n",
      "Epoch 00008: val_loss improved from 0.52693 to 0.49122, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.5023 - acc: 0.7323 - val_loss: 0.4912 - val_acc: 0.7888\n",
      "Epoch 9/150\n",
      "Epoch 00009: val_loss improved from 0.49122 to 0.47260, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.4889 - acc: 0.7506 - val_loss: 0.4726 - val_acc: 0.7826\n",
      "Epoch 10/150\n",
      "Epoch 00010: val_loss improved from 0.47260 to 0.43921, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 13s - loss: 0.4688 - acc: 0.7758 - val_loss: 0.4392 - val_acc: 0.8137\n",
      "Epoch 11/150\n",
      "Epoch 00011: val_loss improved from 0.43921 to 0.43244, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.4646 - acc: 0.7646 - val_loss: 0.4324 - val_acc: 0.8075\n",
      "Epoch 12/150\n",
      "Epoch 00012: val_loss did not improve\n",
      " - 12s - loss: 0.4577 - acc: 0.7880 - val_loss: 0.4549 - val_acc: 0.7857\n",
      "Epoch 13/150\n",
      "Epoch 00013: val_loss did not improve\n",
      " - 12s - loss: 0.4212 - acc: 0.8063 - val_loss: 0.4413 - val_acc: 0.8012\n",
      "Epoch 14/150\n",
      "Epoch 00014: val_loss improved from 0.43244 to 0.39107, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.3957 - acc: 0.8223 - val_loss: 0.3911 - val_acc: 0.8230\n",
      "Epoch 15/150\n",
      "Epoch 00015: val_loss did not improve\n",
      " - 12s - loss: 0.4071 - acc: 0.8109 - val_loss: 0.4523 - val_acc: 0.7919\n",
      "Epoch 16/150\n",
      "Epoch 00016: val_loss did not improve\n",
      " - 12s - loss: 0.3857 - acc: 0.8193 - val_loss: 0.4195 - val_acc: 0.8075\n",
      "Epoch 17/150\n",
      "Epoch 00017: val_loss did not improve\n",
      " - 12s - loss: 0.3862 - acc: 0.8322 - val_loss: 0.4243 - val_acc: 0.7857\n",
      "Epoch 18/150\n",
      "Epoch 00018: val_loss did not improve\n",
      " - 12s - loss: 0.3695 - acc: 0.8168 - val_loss: 0.4224 - val_acc: 0.7981\n",
      "Epoch 19/150\n",
      "Epoch 00019: val_loss improved from 0.39107 to 0.37584, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 13s - loss: 0.3686 - acc: 0.8368 - val_loss: 0.3758 - val_acc: 0.8043\n",
      "Epoch 20/150\n",
      "Epoch 00020: val_loss did not improve\n",
      " - 12s - loss: 0.3540 - acc: 0.8345 - val_loss: 0.4041 - val_acc: 0.8012\n",
      "Epoch 21/150\n",
      "Epoch 00021: val_loss improved from 0.37584 to 0.35188, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.3573 - acc: 0.8353 - val_loss: 0.3519 - val_acc: 0.8261\n",
      "Epoch 22/150\n",
      "Epoch 00022: val_loss improved from 0.35188 to 0.34263, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.3374 - acc: 0.8460 - val_loss: 0.3426 - val_acc: 0.8354\n",
      "Epoch 23/150\n",
      "Epoch 00023: val_loss improved from 0.34263 to 0.33506, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.3568 - acc: 0.8437 - val_loss: 0.3351 - val_acc: 0.8571\n",
      "Epoch 24/150\n",
      "Epoch 00024: val_loss did not improve\n",
      " - 13s - loss: 0.3405 - acc: 0.8348 - val_loss: 0.3635 - val_acc: 0.8292\n",
      "Epoch 25/150\n",
      "Epoch 00025: val_loss did not improve\n",
      " - 12s - loss: 0.3534 - acc: 0.8353 - val_loss: 0.3591 - val_acc: 0.8137\n",
      "Epoch 26/150\n",
      "Epoch 00026: val_loss did not improve\n",
      " - 12s - loss: 0.3161 - acc: 0.8447 - val_loss: 0.7115 - val_acc: 0.6553\n",
      "Epoch 27/150\n",
      "Epoch 00027: val_loss did not improve\n",
      " - 12s - loss: 0.3816 - acc: 0.8208 - val_loss: 0.3784 - val_acc: 0.8012\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 28/150\n",
      "Epoch 00028: val_loss did not improve\n",
      " - 12s - loss: 0.3270 - acc: 0.8376 - val_loss: 0.3811 - val_acc: 0.7981\n",
      "Epoch 29/150\n",
      "Epoch 00029: val_loss improved from 0.33506 to 0.32480, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.3467 - acc: 0.8462 - val_loss: 0.3248 - val_acc: 0.8385\n",
      "Epoch 30/150\n",
      "Epoch 00030: val_loss did not improve\n",
      " - 12s - loss: 0.3258 - acc: 0.8498 - val_loss: 0.3616 - val_acc: 0.7981\n",
      "Epoch 31/150\n",
      "Epoch 00031: val_loss did not improve\n",
      " - 12s - loss: 0.3016 - acc: 0.8665 - val_loss: 0.3854 - val_acc: 0.8168\n",
      "Epoch 32/150\n",
      "Epoch 00032: val_loss did not improve\n",
      " - 12s - loss: 0.3219 - acc: 0.8516 - val_loss: 0.3438 - val_acc: 0.8230\n",
      "Epoch 33/150\n",
      "Epoch 00033: val_loss improved from 0.32480 to 0.29823, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.3209 - acc: 0.8513 - val_loss: 0.2982 - val_acc: 0.8571\n",
      "Epoch 34/150\n",
      "Epoch 00034: val_loss improved from 0.29823 to 0.29575, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.2799 - acc: 0.8704 - val_loss: 0.2957 - val_acc: 0.8478\n",
      "Epoch 35/150\n",
      "Epoch 00035: val_loss did not improve\n",
      " - 12s - loss: 0.3123 - acc: 0.8645 - val_loss: 0.3744 - val_acc: 0.7857\n",
      "Epoch 36/150\n",
      "Epoch 00036: val_loss improved from 0.29575 to 0.29202, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.3176 - acc: 0.8467 - val_loss: 0.2920 - val_acc: 0.8665\n",
      "Epoch 37/150\n",
      "Epoch 00037: val_loss did not improve\n",
      " - 12s - loss: 0.2811 - acc: 0.8691 - val_loss: 0.3249 - val_acc: 0.8354\n",
      "Epoch 38/150\n",
      "Epoch 00038: val_loss did not improve\n",
      " - 12s - loss: 0.2867 - acc: 0.8696 - val_loss: 0.3118 - val_acc: 0.8292\n",
      "Epoch 39/150\n",
      "Epoch 00039: val_loss did not improve\n",
      " - 12s - loss: 0.3090 - acc: 0.8554 - val_loss: 0.3351 - val_acc: 0.8354\n",
      "Epoch 40/150\n",
      "Epoch 00040: val_loss did not improve\n",
      " - 12s - loss: 0.3174 - acc: 0.8627 - val_loss: 0.3209 - val_acc: 0.8323\n",
      "Epoch 41/150\n",
      "Epoch 00041: val_loss improved from 0.29202 to 0.28263, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.2761 - acc: 0.8658 - val_loss: 0.2826 - val_acc: 0.8665\n",
      "Epoch 42/150\n",
      "Epoch 00042: val_loss did not improve\n",
      " - 12s - loss: 0.2675 - acc: 0.8643 - val_loss: 0.3695 - val_acc: 0.8137\n",
      "Epoch 43/150\n",
      "Epoch 00043: val_loss improved from 0.28263 to 0.27834, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.2785 - acc: 0.8765 - val_loss: 0.2783 - val_acc: 0.8447\n",
      "Epoch 44/150\n",
      "Epoch 00044: val_loss did not improve\n",
      " - 12s - loss: 0.2883 - acc: 0.8630 - val_loss: 0.3277 - val_acc: 0.8199\n",
      "Epoch 45/150\n",
      "Epoch 00045: val_loss did not improve\n",
      " - 12s - loss: 0.2774 - acc: 0.8711 - val_loss: 0.2884 - val_acc: 0.8292\n",
      "Epoch 46/150\n",
      "Epoch 00046: val_loss did not improve\n",
      " - 12s - loss: 0.2565 - acc: 0.8887 - val_loss: 0.2960 - val_acc: 0.8416\n",
      "Epoch 47/150\n",
      "Epoch 00047: val_loss improved from 0.27834 to 0.27224, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.2607 - acc: 0.8864 - val_loss: 0.2722 - val_acc: 0.8727\n",
      "Epoch 48/150\n",
      "Epoch 00048: val_loss did not improve\n",
      " - 12s - loss: 0.2469 - acc: 0.8833 - val_loss: 0.2743 - val_acc: 0.8665\n",
      "Epoch 49/150\n",
      "Epoch 00049: val_loss did not improve\n",
      " - 12s - loss: 0.2460 - acc: 0.8909 - val_loss: 0.2747 - val_acc: 0.8571\n",
      "Epoch 50/150\n",
      "Epoch 00050: val_loss did not improve\n",
      " - 12s - loss: 0.2372 - acc: 0.8887 - val_loss: 0.3031 - val_acc: 0.8261\n",
      "Epoch 51/150\n",
      "Epoch 00051: val_loss did not improve\n",
      " - 12s - loss: 0.2357 - acc: 0.9001 - val_loss: 0.2843 - val_acc: 0.8385\n",
      "Epoch 52/150\n",
      "Epoch 00052: val_loss improved from 0.27224 to 0.26110, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.2443 - acc: 0.8925 - val_loss: 0.2611 - val_acc: 0.8665\n",
      "Epoch 53/150\n",
      "Epoch 00053: val_loss did not improve\n",
      " - 12s - loss: 0.2421 - acc: 0.8963 - val_loss: 0.2656 - val_acc: 0.8602\n",
      "Epoch 54/150\n",
      "Epoch 00054: val_loss improved from 0.26110 to 0.25691, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.2243 - acc: 0.9054 - val_loss: 0.2569 - val_acc: 0.8696\n",
      "Epoch 55/150\n",
      "Epoch 00055: val_loss did not improve\n",
      " - 12s - loss: 0.2348 - acc: 0.8894 - val_loss: 0.2633 - val_acc: 0.8602\n",
      "Epoch 56/150\n",
      "Epoch 00056: val_loss did not improve\n",
      " - 12s - loss: 0.2338 - acc: 0.9001 - val_loss: 0.3329 - val_acc: 0.8416\n",
      "Epoch 57/150\n",
      "Epoch 00057: val_loss did not improve\n",
      " - 12s - loss: 0.2923 - acc: 0.8577 - val_loss: 0.2656 - val_acc: 0.8571\n",
      "Epoch 58/150\n",
      "Epoch 00058: val_loss did not improve\n",
      " - 12s - loss: 0.2501 - acc: 0.8795 - val_loss: 0.2777 - val_acc: 0.8540\n",
      "Epoch 59/150\n",
      "Epoch 00059: val_loss did not improve\n",
      " - 12s - loss: 0.2356 - acc: 0.8948 - val_loss: 0.3285 - val_acc: 0.8416\n",
      "Epoch 60/150\n",
      "Epoch 00060: val_loss improved from 0.25691 to 0.25600, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.2178 - acc: 0.8940 - val_loss: 0.2560 - val_acc: 0.8789\n",
      "Epoch 61/150\n",
      "Epoch 00061: val_loss did not improve\n",
      " - 12s - loss: 0.2550 - acc: 0.8767 - val_loss: 0.2851 - val_acc: 0.8385\n",
      "Epoch 62/150\n",
      "Epoch 00062: val_loss did not improve\n",
      " - 12s - loss: 0.2091 - acc: 0.9085 - val_loss: 0.2762 - val_acc: 0.8696\n",
      "Epoch 63/150\n",
      "Epoch 00063: val_loss did not improve\n",
      " - 12s - loss: 0.2720 - acc: 0.8691 - val_loss: 0.2904 - val_acc: 0.8354\n",
      "Epoch 64/150\n",
      "Epoch 00064: val_loss did not improve\n",
      " - 12s - loss: 0.2840 - acc: 0.8622 - val_loss: 0.3340 - val_acc: 0.8634\n",
      "Epoch 65/150\n",
      "Epoch 00065: val_loss did not improve\n",
      " - 12s - loss: 0.2579 - acc: 0.8833 - val_loss: 0.2826 - val_acc: 0.8447\n",
      "Epoch 66/150\n",
      "Epoch 00066: val_loss did not improve\n",
      " - 12s - loss: 0.2382 - acc: 0.8966 - val_loss: 0.2701 - val_acc: 0.8665\n",
      "Epoch 67/150\n",
      "Epoch 00067: val_loss did not improve\n",
      " - 12s - loss: 0.2131 - acc: 0.9070 - val_loss: 0.2714 - val_acc: 0.8665\n",
      "Epoch 68/150\n",
      "Epoch 00068: val_loss did not improve\n",
      " - 12s - loss: 0.2166 - acc: 0.9146 - val_loss: 0.2731 - val_acc: 0.8634\n",
      "Epoch 69/150\n",
      "Epoch 00069: val_loss did not improve\n",
      " - 12s - loss: 0.2404 - acc: 0.8993 - val_loss: 0.2664 - val_acc: 0.8634\n",
      "Epoch 70/150\n",
      "Epoch 00070: val_loss did not improve\n",
      " - 12s - loss: 0.2293 - acc: 0.9070 - val_loss: 0.2584 - val_acc: 0.8758\n",
      "Epoch 71/150\n",
      "Epoch 00071: val_loss did not improve\n",
      " - 12s - loss: 0.2303 - acc: 0.8935 - val_loss: 0.2850 - val_acc: 0.8665\n",
      "Epoch 72/150\n",
      "Epoch 00072: val_loss improved from 0.25600 to 0.25320, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.2376 - acc: 0.8948 - val_loss: 0.2532 - val_acc: 0.8727\n",
      "Epoch 73/150\n",
      "Epoch 00073: val_loss did not improve\n",
      " - 12s - loss: 0.2205 - acc: 0.9047 - val_loss: 0.2557 - val_acc: 0.8696\n",
      "Epoch 74/150\n",
      "Epoch 00074: val_loss did not improve\n",
      " - 12s - loss: 0.2522 - acc: 0.8790 - val_loss: 0.2978 - val_acc: 0.8354\n",
      "Epoch 75/150\n",
      "Epoch 00075: val_loss did not improve\n",
      " - 12s - loss: 0.2291 - acc: 0.8970 - val_loss: 0.2780 - val_acc: 0.8602\n",
      "Epoch 76/150\n",
      "Epoch 00076: val_loss did not improve\n",
      " - 12s - loss: 0.2194 - acc: 0.9009 - val_loss: 0.3337 - val_acc: 0.8168\n",
      "Epoch 77/150\n",
      "Epoch 00077: val_loss did not improve\n",
      " - 12s - loss: 0.2081 - acc: 0.9077 - val_loss: 0.2613 - val_acc: 0.8696\n",
      "Epoch 78/150\n",
      "Epoch 00078: val_loss did not improve\n",
      " - 12s - loss: 0.2111 - acc: 0.9092 - val_loss: 0.2549 - val_acc: 0.8696\n",
      "Epoch 79/150\n",
      "Epoch 00079: val_loss did not improve\n",
      " - 12s - loss: 0.2057 - acc: 0.9184 - val_loss: 0.2673 - val_acc: 0.8634\n",
      "Epoch 80/150\n",
      "Epoch 00080: val_loss did not improve\n",
      " - 12s - loss: 0.2382 - acc: 0.8912 - val_loss: 0.2872 - val_acc: 0.8323\n",
      "Epoch 81/150\n",
      "Epoch 00081: val_loss did not improve\n",
      " - 12s - loss: 0.2358 - acc: 0.9039 - val_loss: 0.2708 - val_acc: 0.8509\n",
      "Epoch 82/150\n",
      "Epoch 00082: val_loss improved from 0.25320 to 0.24297, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv1.hdf5\n",
      " - 12s - loss: 0.2139 - acc: 0.9108 - val_loss: 0.2430 - val_acc: 0.8820\n",
      "Epoch 83/150\n",
      "Epoch 00083: val_loss did not improve\n",
      " - 12s - loss: 0.1985 - acc: 0.9153 - val_loss: 0.2577 - val_acc: 0.8665\n",
      "Epoch 84/150\n",
      "Epoch 00084: val_loss did not improve\n",
      " - 12s - loss: 0.2155 - acc: 0.9027 - val_loss: 0.2811 - val_acc: 0.8571\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 85/150\n",
      "Epoch 00085: val_loss did not improve\n",
      " - 12s - loss: 0.2729 - acc: 0.8706 - val_loss: 0.3325 - val_acc: 0.8261\n",
      "Epoch 86/150\n",
      "Epoch 00086: val_loss did not improve\n",
      " - 12s - loss: 0.2348 - acc: 0.9024 - val_loss: 0.2898 - val_acc: 0.8602\n",
      "Epoch 87/150\n",
      "Epoch 00087: val_loss did not improve\n",
      " - 12s - loss: 0.2246 - acc: 0.8963 - val_loss: 0.2620 - val_acc: 0.8602\n",
      "Epoch 88/150\n",
      "Epoch 00088: val_loss did not improve\n",
      " - 12s - loss: 0.2598 - acc: 0.8813 - val_loss: 0.2777 - val_acc: 0.8447\n",
      "Epoch 89/150\n",
      "Epoch 00089: val_loss did not improve\n",
      " - 12s - loss: 0.2110 - acc: 0.9161 - val_loss: 0.2652 - val_acc: 0.8727\n",
      "Epoch 90/150\n",
      "Epoch 00090: val_loss did not improve\n",
      " - 12s - loss: 0.2091 - acc: 0.9092 - val_loss: 0.2477 - val_acc: 0.8944\n",
      "Epoch 91/150\n",
      "Epoch 00091: val_loss did not improve\n",
      " - 12s - loss: 0.1985 - acc: 0.9199 - val_loss: 0.2587 - val_acc: 0.8820\n",
      "Epoch 92/150\n",
      "Epoch 00092: val_loss did not improve\n",
      " - 12s - loss: 0.1921 - acc: 0.9260 - val_loss: 0.2697 - val_acc: 0.8665\n",
      "Epoch 93/150\n",
      "Epoch 00093: val_loss did not improve\n",
      " - 12s - loss: 0.2553 - acc: 0.8767 - val_loss: 0.3205 - val_acc: 0.8261\n",
      "Epoch 94/150\n",
      "Epoch 00094: val_loss did not improve\n",
      " - 12s - loss: 0.2108 - acc: 0.9031 - val_loss: 0.2945 - val_acc: 0.8540\n",
      "Epoch 95/150\n",
      "Epoch 00095: val_loss did not improve\n",
      " - 12s - loss: 0.1949 - acc: 0.9146 - val_loss: 0.2492 - val_acc: 0.8882\n",
      "Epoch 96/150\n",
      "Epoch 00096: val_loss did not improve\n",
      " - 12s - loss: 0.1939 - acc: 0.9169 - val_loss: 0.2571 - val_acc: 0.8789\n",
      "Epoch 97/150\n",
      "Epoch 00097: val_loss did not improve\n",
      " - 12s - loss: 0.2005 - acc: 0.8989 - val_loss: 0.2505 - val_acc: 0.8851\n",
      "Epoch 98/150\n",
      "Epoch 00098: val_loss did not improve\n",
      " - 12s - loss: 0.1987 - acc: 0.9085 - val_loss: 0.2779 - val_acc: 0.8571\n",
      "Epoch 99/150\n",
      "Epoch 00099: val_loss did not improve\n",
      " - 12s - loss: 0.1929 - acc: 0.9169 - val_loss: 0.2759 - val_acc: 0.8789\n",
      "Epoch 100/150\n",
      "Epoch 00100: val_loss did not improve\n",
      " - 12s - loss: 0.1896 - acc: 0.9199 - val_loss: 0.2490 - val_acc: 0.8882\n",
      "Epoch 101/150\n",
      "Epoch 00101: val_loss did not improve\n",
      " - 12s - loss: 0.1764 - acc: 0.9276 - val_loss: 0.2446 - val_acc: 0.8882\n",
      "Epoch 102/150\n",
      "Epoch 00102: val_loss did not improve\n",
      " - 12s - loss: 0.2232 - acc: 0.8973 - val_loss: 0.2603 - val_acc: 0.8820\n",
      "Epoch 103/150\n",
      "Epoch 00103: val_loss did not improve\n",
      " - 12s - loss: 0.1852 - acc: 0.9230 - val_loss: 0.2759 - val_acc: 0.8634\n",
      "Epoch 104/150\n",
      "Epoch 00104: val_loss did not improve\n",
      " - 12s - loss: 0.2064 - acc: 0.8993 - val_loss: 0.2852 - val_acc: 0.8602\n",
      "Epoch 105/150\n",
      "Epoch 00105: val_loss did not improve\n",
      " - 12s - loss: 0.1760 - acc: 0.9268 - val_loss: 0.2645 - val_acc: 0.8851\n",
      "Epoch 106/150\n",
      "Epoch 00106: val_loss did not improve\n",
      " - 12s - loss: 0.1895 - acc: 0.9146 - val_loss: 0.2823 - val_acc: 0.8447\n",
      "Epoch 107/150\n",
      "Epoch 00107: val_loss did not improve\n",
      " - 12s - loss: 0.1823 - acc: 0.9123 - val_loss: 0.2578 - val_acc: 0.8851\n",
      "Epoch 108/150\n",
      "Epoch 00108: val_loss did not improve\n",
      " - 12s - loss: 0.1787 - acc: 0.9268 - val_loss: 0.2513 - val_acc: 0.8851\n",
      "Epoch 109/150\n",
      "Epoch 00109: val_loss did not improve\n",
      " - 12s - loss: 0.1721 - acc: 0.9253 - val_loss: 0.2466 - val_acc: 0.8944\n",
      "Epoch 110/150\n",
      "Epoch 00110: val_loss did not improve\n",
      " - 12s - loss: 0.1669 - acc: 0.9344 - val_loss: 0.2696 - val_acc: 0.8851\n",
      "Epoch 111/150\n",
      "Epoch 00111: val_loss did not improve\n",
      " - 12s - loss: 0.2177 - acc: 0.8950 - val_loss: 0.2743 - val_acc: 0.8665\n",
      "Epoch 112/150\n",
      "Epoch 00112: val_loss did not improve\n",
      " - 12s - loss: 0.1638 - acc: 0.9375 - val_loss: 0.2524 - val_acc: 0.8882\n",
      "Satge 2,done!\n",
      "1282/1282 [==============================] - 3s 3ms/step\n",
      "Train loss: 0.180572447557\n",
      "Train accuracy: 0.92511700468\n",
      "322/322 [==============================] - 1s 3ms/step\n",
      "Valid loss: 0.242973468318\n",
      "Valid accuracy: 0.88198757764\n",
      "\n",
      "===================FOLD= 2\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_2 (InputLayer)         (None, 75, 75, 3)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_5 (Conv2D)            (None, 71, 71, 128)       9728      \n",
      "_________________________________________________________________\n",
      "activation_5 (Activation)    (None, 71, 71, 128)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_5 (MaxPooling2 (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "dropout_7 (Dropout)          (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_6 (Conv2D)            (None, 33, 33, 256)       295168    \n",
      "_________________________________________________________________\n",
      "activation_6 (Activation)    (None, 33, 33, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_6 (MaxPooling2 (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "dropout_8 (Dropout)          (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_7 (Conv2D)            (None, 14, 14, 256)       590080    \n",
      "_________________________________________________________________\n",
      "activation_7 (Activation)    (None, 14, 14, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_7 (MaxPooling2 (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "dropout_9 (Dropout)          (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_8 (Conv2D)            (None, 5, 5, 128)         295040    \n",
      "_________________________________________________________________\n",
      "activation_8 (Activation)    (None, 5, 5, 128)         0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_8 (MaxPooling2 (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "dropout_10 (Dropout)         (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten_2 (Flatten)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dropout_11 (Dropout)         (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 256)               131328    \n",
      "_________________________________________________________________\n",
      "dropout_12 (Dropout)         (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_6 (Dense)              (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 1,584,257\n",
      "Trainable params: 1,584,257\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/150\n",
      "Epoch 00001: val_loss improved from inf to 0.69481, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 15s - loss: 1.3244 - acc: 0.4975 - val_loss: 0.6948 - val_acc: 0.4642\n",
      "Epoch 2/150\n",
      "Epoch 00002: val_loss improved from 0.69481 to 0.68431, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.7500 - acc: 0.5082 - val_loss: 0.6843 - val_acc: 0.5888\n",
      "Epoch 3/150\n",
      "Epoch 00003: val_loss improved from 0.68431 to 0.59997, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.6845 - acc: 0.5643 - val_loss: 0.6000 - val_acc: 0.6978\n",
      "Epoch 4/150\n",
      "Epoch 00004: val_loss improved from 0.59997 to 0.55580, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.6250 - acc: 0.5973 - val_loss: 0.5558 - val_acc: 0.7009\n",
      "Epoch 5/150\n",
      "Epoch 00005: val_loss improved from 0.55580 to 0.54407, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.5944 - acc: 0.6074 - val_loss: 0.5441 - val_acc: 0.7072\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 6/150\n",
      "Epoch 00006: val_loss improved from 0.54407 to 0.53886, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.5729 - acc: 0.6580 - val_loss: 0.5389 - val_acc: 0.7227\n",
      "Epoch 7/150\n",
      "Epoch 00007: val_loss improved from 0.53886 to 0.53091, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.5714 - acc: 0.6226 - val_loss: 0.5309 - val_acc: 0.7352\n",
      "Epoch 8/150\n",
      "Epoch 00008: val_loss improved from 0.53091 to 0.52050, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.5479 - acc: 0.6733 - val_loss: 0.5205 - val_acc: 0.7882\n",
      "Epoch 9/150\n",
      "Epoch 00009: val_loss improved from 0.52050 to 0.48095, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.5372 - acc: 0.6943 - val_loss: 0.4810 - val_acc: 0.7726\n",
      "Epoch 10/150\n",
      "Epoch 00010: val_loss improved from 0.48095 to 0.46453, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.5353 - acc: 0.7175 - val_loss: 0.4645 - val_acc: 0.8006\n",
      "Epoch 11/150\n",
      "Epoch 00011: val_loss improved from 0.46453 to 0.44848, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.4875 - acc: 0.7628 - val_loss: 0.4485 - val_acc: 0.7757\n",
      "Epoch 12/150\n",
      "Epoch 00012: val_loss improved from 0.44848 to 0.38748, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.4459 - acc: 0.7922 - val_loss: 0.3875 - val_acc: 0.8255\n",
      "Epoch 13/150\n",
      "Epoch 00013: val_loss improved from 0.38748 to 0.38728, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.4289 - acc: 0.7866 - val_loss: 0.3873 - val_acc: 0.8318\n",
      "Epoch 14/150\n",
      "Epoch 00014: val_loss did not improve\n",
      " - 12s - loss: 0.4502 - acc: 0.7884 - val_loss: 0.5300 - val_acc: 0.7134\n",
      "Epoch 15/150\n",
      "Epoch 00015: val_loss did not improve\n",
      " - 12s - loss: 0.4816 - acc: 0.7843 - val_loss: 0.4193 - val_acc: 0.8349\n",
      "Epoch 16/150\n",
      "Epoch 00016: val_loss improved from 0.38728 to 0.33719, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.4192 - acc: 0.8162 - val_loss: 0.3372 - val_acc: 0.8629\n",
      "Epoch 17/150\n",
      "Epoch 00017: val_loss did not improve\n",
      " - 12s - loss: 0.4048 - acc: 0.8022 - val_loss: 0.3533 - val_acc: 0.8598\n",
      "Epoch 18/150\n",
      "Epoch 00018: val_loss did not improve\n",
      " - 12s - loss: 0.3907 - acc: 0.8124 - val_loss: 0.3641 - val_acc: 0.8380\n",
      "Epoch 19/150\n",
      "Epoch 00019: val_loss improved from 0.33719 to 0.29089, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.3774 - acc: 0.8216 - val_loss: 0.2909 - val_acc: 0.8692\n",
      "Epoch 20/150\n",
      "Epoch 00020: val_loss improved from 0.29089 to 0.28063, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.3712 - acc: 0.8322 - val_loss: 0.2806 - val_acc: 0.8816\n",
      "Epoch 21/150\n",
      "Epoch 00021: val_loss did not improve\n",
      " - 12s - loss: 0.3427 - acc: 0.8475 - val_loss: 0.2823 - val_acc: 0.8692\n",
      "Epoch 22/150\n",
      "Epoch 00022: val_loss improved from 0.28063 to 0.27555, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.3628 - acc: 0.8345 - val_loss: 0.2755 - val_acc: 0.8972\n",
      "Epoch 23/150\n",
      "Epoch 00023: val_loss did not improve\n",
      " - 12s - loss: 0.3394 - acc: 0.8399 - val_loss: 0.3039 - val_acc: 0.8629\n",
      "Epoch 24/150\n",
      "Epoch 00024: val_loss improved from 0.27555 to 0.26822, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.3427 - acc: 0.8327 - val_loss: 0.2682 - val_acc: 0.8692\n",
      "Epoch 25/150\n",
      "Epoch 00025: val_loss did not improve\n",
      " - 12s - loss: 0.3339 - acc: 0.8482 - val_loss: 0.3565 - val_acc: 0.8037\n",
      "Epoch 26/150\n",
      "Epoch 00026: val_loss improved from 0.26822 to 0.25154, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.3050 - acc: 0.8620 - val_loss: 0.2515 - val_acc: 0.8910\n",
      "Epoch 27/150\n",
      "Epoch 00027: val_loss improved from 0.25154 to 0.23722, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.3063 - acc: 0.8612 - val_loss: 0.2372 - val_acc: 0.8972\n",
      "Epoch 28/150\n",
      "Epoch 00028: val_loss did not improve\n",
      " - 12s - loss: 0.3326 - acc: 0.8342 - val_loss: 0.2549 - val_acc: 0.8910\n",
      "Epoch 29/150\n",
      "Epoch 00029: val_loss did not improve\n",
      " - 12s - loss: 0.3078 - acc: 0.8536 - val_loss: 0.2557 - val_acc: 0.8972\n",
      "Epoch 30/150\n",
      "Epoch 00030: val_loss did not improve\n",
      " - 12s - loss: 0.3156 - acc: 0.8563 - val_loss: 0.2378 - val_acc: 0.8910\n",
      "Epoch 31/150\n",
      "Epoch 00031: val_loss did not improve\n",
      " - 12s - loss: 0.2905 - acc: 0.8665 - val_loss: 0.2855 - val_acc: 0.8723\n",
      "Epoch 32/150\n",
      "Epoch 00032: val_loss did not improve\n",
      " - 12s - loss: 0.2809 - acc: 0.8566 - val_loss: 0.2633 - val_acc: 0.8629\n",
      "Epoch 33/150\n",
      "Epoch 00033: val_loss did not improve\n",
      " - 12s - loss: 0.3198 - acc: 0.8514 - val_loss: 0.2538 - val_acc: 0.8910\n",
      "Epoch 34/150\n",
      "Epoch 00034: val_loss did not improve\n",
      " - 12s - loss: 0.3136 - acc: 0.8673 - val_loss: 0.2583 - val_acc: 0.8816\n",
      "Epoch 35/150\n",
      "Epoch 00035: val_loss did not improve\n",
      " - 12s - loss: 0.3095 - acc: 0.8533 - val_loss: 0.2378 - val_acc: 0.9003\n",
      "Epoch 36/150\n",
      "Epoch 00036: val_loss did not improve\n",
      " - 12s - loss: 0.2764 - acc: 0.8772 - val_loss: 0.2658 - val_acc: 0.8879\n",
      "Epoch 37/150\n",
      "Epoch 00037: val_loss did not improve\n",
      " - 12s - loss: 0.2782 - acc: 0.8670 - val_loss: 0.2702 - val_acc: 0.8567\n",
      "Epoch 38/150\n",
      "Epoch 00038: val_loss improved from 0.23722 to 0.22474, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.2585 - acc: 0.8787 - val_loss: 0.2247 - val_acc: 0.9065\n",
      "Epoch 39/150\n",
      "Epoch 00039: val_loss did not improve\n",
      " - 12s - loss: 0.2684 - acc: 0.8841 - val_loss: 0.3139 - val_acc: 0.8474\n",
      "Epoch 40/150\n",
      "Epoch 00040: val_loss improved from 0.22474 to 0.22043, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.2777 - acc: 0.8659 - val_loss: 0.2204 - val_acc: 0.8972\n",
      "Epoch 41/150\n",
      "Epoch 00041: val_loss did not improve\n",
      " - 12s - loss: 0.2550 - acc: 0.8826 - val_loss: 0.2594 - val_acc: 0.8629\n",
      "Epoch 42/150\n",
      "Epoch 00042: val_loss improved from 0.22043 to 0.21097, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.2440 - acc: 0.8940 - val_loss: 0.2110 - val_acc: 0.9128\n",
      "Epoch 43/150\n",
      "Epoch 00043: val_loss did not improve\n",
      " - 12s - loss: 0.2416 - acc: 0.9016 - val_loss: 0.2162 - val_acc: 0.9065\n",
      "Epoch 44/150\n",
      "Epoch 00044: val_loss did not improve\n",
      " - 12s - loss: 0.2718 - acc: 0.8818 - val_loss: 0.2628 - val_acc: 0.8692\n",
      "Epoch 45/150\n",
      "Epoch 00045: val_loss did not improve\n",
      " - 12s - loss: 0.2524 - acc: 0.8955 - val_loss: 0.2887 - val_acc: 0.8660\n",
      "Epoch 46/150\n",
      "Epoch 00046: val_loss did not improve\n",
      " - 12s - loss: 0.2809 - acc: 0.8704 - val_loss: 0.2669 - val_acc: 0.8660\n",
      "Epoch 47/150\n",
      "Epoch 00047: val_loss improved from 0.21097 to 0.21042, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.2541 - acc: 0.8963 - val_loss: 0.2104 - val_acc: 0.9159\n",
      "Epoch 48/150\n",
      "Epoch 00048: val_loss did not improve\n",
      " - 12s - loss: 0.2430 - acc: 0.8864 - val_loss: 0.2138 - val_acc: 0.9159\n",
      "Epoch 49/150\n",
      "Epoch 00049: val_loss did not improve\n",
      " - 12s - loss: 0.2350 - acc: 0.8853 - val_loss: 0.2218 - val_acc: 0.8910\n",
      "Epoch 50/150\n",
      "Epoch 00050: val_loss did not improve\n",
      " - 12s - loss: 0.2493 - acc: 0.8856 - val_loss: 0.2152 - val_acc: 0.9097\n",
      "Epoch 51/150\n",
      "Epoch 00051: val_loss did not improve\n",
      " - 12s - loss: 0.2510 - acc: 0.8871 - val_loss: 0.2163 - val_acc: 0.9097\n",
      "Epoch 52/150\n",
      "Epoch 00052: val_loss did not improve\n",
      " - 12s - loss: 0.2304 - acc: 0.8948 - val_loss: 0.2954 - val_acc: 0.8255\n",
      "Epoch 53/150\n",
      "Epoch 00053: val_loss improved from 0.21042 to 0.20329, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.2449 - acc: 0.8887 - val_loss: 0.2033 - val_acc: 0.9252\n",
      "Epoch 54/150\n",
      "Epoch 00054: val_loss did not improve\n",
      " - 12s - loss: 0.2349 - acc: 0.8970 - val_loss: 0.2112 - val_acc: 0.9159\n",
      "Epoch 55/150\n",
      "Epoch 00055: val_loss did not improve\n",
      " - 12s - loss: 0.2447 - acc: 0.8754 - val_loss: 0.2117 - val_acc: 0.9190\n",
      "Epoch 56/150\n",
      "Epoch 00056: val_loss did not improve\n",
      " - 12s - loss: 0.2726 - acc: 0.8612 - val_loss: 0.2325 - val_acc: 0.8972\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 57/150\n",
      "Epoch 00057: val_loss did not improve\n",
      " - 12s - loss: 0.2482 - acc: 0.8807 - val_loss: 0.2384 - val_acc: 0.8941\n",
      "Epoch 58/150\n",
      "Epoch 00058: val_loss did not improve\n",
      " - 12s - loss: 0.2473 - acc: 0.8871 - val_loss: 0.2431 - val_acc: 0.8910\n",
      "Epoch 59/150\n",
      "Epoch 00059: val_loss did not improve\n",
      " - 12s - loss: 0.2250 - acc: 0.8963 - val_loss: 0.2116 - val_acc: 0.9034\n",
      "Epoch 60/150\n",
      "Epoch 00060: val_loss did not improve\n",
      " - 12s - loss: 0.2501 - acc: 0.8990 - val_loss: 0.3748 - val_acc: 0.8629\n",
      "Epoch 61/150\n",
      "Epoch 00061: val_loss did not improve\n",
      " - 12s - loss: 0.2603 - acc: 0.8826 - val_loss: 0.2057 - val_acc: 0.9252\n",
      "Epoch 62/150\n",
      "Epoch 00062: val_loss did not improve\n",
      " - 12s - loss: 0.2432 - acc: 0.8921 - val_loss: 0.2198 - val_acc: 0.9034\n",
      "Epoch 63/150\n",
      "Epoch 00063: val_loss did not improve\n",
      " - 12s - loss: 0.2269 - acc: 0.8937 - val_loss: 0.2518 - val_acc: 0.8629\n",
      "Epoch 64/150\n",
      "Epoch 00064: val_loss did not improve\n",
      " - 12s - loss: 0.2175 - acc: 0.8940 - val_loss: 0.2295 - val_acc: 0.9097\n",
      "Epoch 65/150\n",
      "Epoch 00065: val_loss improved from 0.20329 to 0.20028, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.2402 - acc: 0.8940 - val_loss: 0.2003 - val_acc: 0.9283\n",
      "Epoch 66/150\n",
      "Epoch 00066: val_loss did not improve\n",
      " - 12s - loss: 0.2151 - acc: 0.9100 - val_loss: 0.2371 - val_acc: 0.8847\n",
      "Epoch 67/150\n",
      "Epoch 00067: val_loss did not improve\n",
      " - 12s - loss: 0.2194 - acc: 0.8975 - val_loss: 0.2099 - val_acc: 0.9097\n",
      "Epoch 68/150\n",
      "Epoch 00068: val_loss did not improve\n",
      " - 12s - loss: 0.2242 - acc: 0.8970 - val_loss: 0.2013 - val_acc: 0.9190\n",
      "Epoch 69/150\n",
      "Epoch 00069: val_loss did not improve\n",
      " - 12s - loss: 0.2243 - acc: 0.8993 - val_loss: 0.2277 - val_acc: 0.9128\n",
      "Epoch 70/150\n",
      "Epoch 00070: val_loss did not improve\n",
      " - 12s - loss: 0.2138 - acc: 0.9100 - val_loss: 0.2264 - val_acc: 0.9065\n",
      "Epoch 71/150\n",
      "Epoch 00071: val_loss did not improve\n",
      " - 12s - loss: 0.2118 - acc: 0.8963 - val_loss: 0.2118 - val_acc: 0.9065\n",
      "Epoch 72/150\n",
      "Epoch 00072: val_loss did not improve\n",
      " - 12s - loss: 0.2567 - acc: 0.8777 - val_loss: 0.2351 - val_acc: 0.8972\n",
      "Epoch 73/150\n",
      "Epoch 00073: val_loss improved from 0.20028 to 0.19422, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.2130 - acc: 0.9077 - val_loss: 0.1942 - val_acc: 0.9283\n",
      "Epoch 74/150\n",
      "Epoch 00074: val_loss did not improve\n",
      " - 12s - loss: 0.2104 - acc: 0.9016 - val_loss: 0.2253 - val_acc: 0.9065\n",
      "Epoch 75/150\n",
      "Epoch 00075: val_loss improved from 0.19422 to 0.19022, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.2135 - acc: 0.9001 - val_loss: 0.1902 - val_acc: 0.9190\n",
      "Epoch 76/150\n",
      "Epoch 00076: val_loss did not improve\n",
      " - 12s - loss: 0.2013 - acc: 0.9123 - val_loss: 0.1941 - val_acc: 0.9221\n",
      "Epoch 77/150\n",
      "Epoch 00077: val_loss did not improve\n",
      " - 12s - loss: 0.2016 - acc: 0.9138 - val_loss: 0.1994 - val_acc: 0.9128\n",
      "Epoch 78/150\n",
      "Epoch 00078: val_loss did not improve\n",
      " - 12s - loss: 0.1970 - acc: 0.9154 - val_loss: 0.2042 - val_acc: 0.9097\n",
      "Epoch 79/150\n",
      "Epoch 00079: val_loss did not improve\n",
      " - 12s - loss: 0.2063 - acc: 0.9077 - val_loss: 0.2091 - val_acc: 0.9034\n",
      "Epoch 80/150\n",
      "Epoch 00080: val_loss did not improve\n",
      " - 12s - loss: 0.2061 - acc: 0.9123 - val_loss: 0.2034 - val_acc: 0.9065\n",
      "Epoch 81/150\n",
      "Epoch 00081: val_loss did not improve\n",
      " - 12s - loss: 0.2019 - acc: 0.9108 - val_loss: 0.2146 - val_acc: 0.9003\n",
      "Epoch 82/150\n",
      "Epoch 00082: val_loss did not improve\n",
      " - 12s - loss: 0.1964 - acc: 0.9146 - val_loss: 0.1954 - val_acc: 0.9190\n",
      "Epoch 83/150\n",
      "Epoch 00083: val_loss did not improve\n",
      " - 12s - loss: 0.2111 - acc: 0.9123 - val_loss: 0.2333 - val_acc: 0.8941\n",
      "Epoch 84/150\n",
      "Epoch 00084: val_loss did not improve\n",
      " - 12s - loss: 0.2036 - acc: 0.9123 - val_loss: 0.2266 - val_acc: 0.8847\n",
      "Epoch 85/150\n",
      "Epoch 00085: val_loss did not improve\n",
      " - 12s - loss: 0.1972 - acc: 0.9184 - val_loss: 0.2113 - val_acc: 0.9034\n",
      "Epoch 86/150\n",
      "Epoch 00086: val_loss did not improve\n",
      " - 12s - loss: 0.1880 - acc: 0.9222 - val_loss: 0.2023 - val_acc: 0.9128\n",
      "Epoch 87/150\n",
      "Epoch 00087: val_loss did not improve\n",
      " - 12s - loss: 0.1936 - acc: 0.9176 - val_loss: 0.2024 - val_acc: 0.9065\n",
      "Epoch 88/150\n",
      "Epoch 00088: val_loss did not improve\n",
      " - 12s - loss: 0.2026 - acc: 0.9123 - val_loss: 0.1920 - val_acc: 0.9315\n",
      "Epoch 89/150\n",
      "Epoch 00089: val_loss did not improve\n",
      " - 12s - loss: 0.2025 - acc: 0.9176 - val_loss: 0.2046 - val_acc: 0.9159\n",
      "Epoch 90/150\n",
      "Epoch 00090: val_loss did not improve\n",
      " - 12s - loss: 0.1866 - acc: 0.9192 - val_loss: 0.1986 - val_acc: 0.9128\n",
      "Epoch 91/150\n",
      "Epoch 00091: val_loss did not improve\n",
      " - 12s - loss: 0.1882 - acc: 0.9253 - val_loss: 0.1952 - val_acc: 0.9190\n",
      "Epoch 92/150\n",
      "Epoch 00092: val_loss did not improve\n",
      " - 12s - loss: 0.2027 - acc: 0.9120 - val_loss: 0.2057 - val_acc: 0.9065\n",
      "Epoch 93/150\n",
      "Epoch 00093: val_loss did not improve\n",
      " - 12s - loss: 0.2007 - acc: 0.9146 - val_loss: 0.2078 - val_acc: 0.9346\n",
      "Epoch 94/150\n",
      "Epoch 00094: val_loss did not improve\n",
      " - 12s - loss: 0.2064 - acc: 0.9108 - val_loss: 0.2305 - val_acc: 0.8847\n",
      "Epoch 95/150\n",
      "Epoch 00095: val_loss did not improve\n",
      " - 12s - loss: 0.1781 - acc: 0.9222 - val_loss: 0.1921 - val_acc: 0.9315\n",
      "Epoch 96/150\n",
      "Epoch 00096: val_loss did not improve\n",
      " - 12s - loss: 0.2051 - acc: 0.9089 - val_loss: 0.2298 - val_acc: 0.8879\n",
      "Epoch 97/150\n",
      "Epoch 00097: val_loss did not improve\n",
      " - 12s - loss: 0.2095 - acc: 0.9001 - val_loss: 0.2210 - val_acc: 0.8910\n",
      "Epoch 98/150\n",
      "Epoch 00098: val_loss improved from 0.19022 to 0.18740, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.1888 - acc: 0.9184 - val_loss: 0.1874 - val_acc: 0.9252\n",
      "Epoch 99/150\n",
      "Epoch 00099: val_loss did not improve\n",
      " - 12s - loss: 0.2053 - acc: 0.9085 - val_loss: 0.2226 - val_acc: 0.9097\n",
      "Epoch 100/150\n",
      "Epoch 00100: val_loss did not improve\n",
      " - 12s - loss: 0.1983 - acc: 0.9021 - val_loss: 0.2026 - val_acc: 0.9221\n",
      "Epoch 101/150\n",
      "Epoch 00101: val_loss did not improve\n",
      " - 12s - loss: 0.1973 - acc: 0.9089 - val_loss: 0.2258 - val_acc: 0.9065\n",
      "Epoch 102/150\n",
      "Epoch 00102: val_loss did not improve\n",
      " - 12s - loss: 0.2532 - acc: 0.8944 - val_loss: 0.2422 - val_acc: 0.9034\n",
      "Epoch 103/150\n",
      "Epoch 00103: val_loss did not improve\n",
      " - 12s - loss: 0.2127 - acc: 0.8944 - val_loss: 0.2070 - val_acc: 0.9034\n",
      "Epoch 104/150\n",
      "Epoch 00104: val_loss improved from 0.18740 to 0.18531, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.1873 - acc: 0.9192 - val_loss: 0.1853 - val_acc: 0.9252\n",
      "Epoch 105/150\n",
      "Epoch 00105: val_loss did not improve\n",
      " - 12s - loss: 0.1981 - acc: 0.9154 - val_loss: 0.1991 - val_acc: 0.9190\n",
      "Epoch 106/150\n",
      "Epoch 00106: val_loss did not improve\n",
      " - 12s - loss: 0.1700 - acc: 0.9298 - val_loss: 0.2016 - val_acc: 0.9097\n",
      "Epoch 107/150\n",
      "Epoch 00107: val_loss did not improve\n",
      " - 12s - loss: 0.2876 - acc: 0.8636 - val_loss: 0.2348 - val_acc: 0.8941\n",
      "Epoch 108/150\n",
      "Epoch 00108: val_loss did not improve\n",
      " - 12s - loss: 0.2362 - acc: 0.8894 - val_loss: 0.1991 - val_acc: 0.9283\n",
      "Epoch 109/150\n",
      "Epoch 00109: val_loss did not improve\n",
      " - 12s - loss: 0.2109 - acc: 0.8929 - val_loss: 0.1971 - val_acc: 0.9346\n",
      "Epoch 110/150\n",
      "Epoch 00110: val_loss did not improve\n",
      " - 12s - loss: 0.1913 - acc: 0.9184 - val_loss: 0.1905 - val_acc: 0.9190\n",
      "Epoch 111/150\n",
      "Epoch 00111: val_loss did not improve\n",
      " - 12s - loss: 0.1703 - acc: 0.9291 - val_loss: 0.1941 - val_acc: 0.9159\n",
      "Epoch 112/150\n",
      "Epoch 00112: val_loss did not improve\n",
      " - 12s - loss: 0.1859 - acc: 0.9154 - val_loss: 0.1998 - val_acc: 0.9128\n",
      "Epoch 113/150\n",
      "Epoch 00113: val_loss did not improve\n",
      " - 12s - loss: 0.1823 - acc: 0.9169 - val_loss: 0.1958 - val_acc: 0.9128\n",
      "Epoch 114/150\n",
      "Epoch 00114: val_loss did not improve\n",
      " - 12s - loss: 0.1859 - acc: 0.9176 - val_loss: 0.1932 - val_acc: 0.9252\n",
      "Epoch 115/150\n",
      "Epoch 00115: val_loss did not improve\n",
      " - 12s - loss: 0.1713 - acc: 0.9222 - val_loss: 0.1921 - val_acc: 0.9252\n",
      "Epoch 116/150\n",
      "Epoch 00116: val_loss did not improve\n",
      " - 12s - loss: 0.1766 - acc: 0.9291 - val_loss: 0.1955 - val_acc: 0.9252\n",
      "Epoch 117/150\n",
      "Epoch 00117: val_loss did not improve\n",
      " - 13s - loss: 0.1796 - acc: 0.9192 - val_loss: 0.2070 - val_acc: 0.9221\n",
      "Epoch 118/150\n",
      "Epoch 00118: val_loss did not improve\n",
      " - 12s - loss: 0.1760 - acc: 0.9222 - val_loss: 0.1940 - val_acc: 0.9252\n",
      "Epoch 119/150\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 00119: val_loss did not improve\n",
      " - 12s - loss: 0.1805 - acc: 0.9207 - val_loss: 0.2056 - val_acc: 0.9128\n",
      "Epoch 120/150\n",
      "Epoch 00120: val_loss did not improve\n",
      " - 12s - loss: 0.2012 - acc: 0.9036 - val_loss: 0.2085 - val_acc: 0.9159\n",
      "Epoch 121/150\n",
      "Epoch 00121: val_loss improved from 0.18531 to 0.17980, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv2.hdf5\n",
      " - 12s - loss: 0.1688 - acc: 0.9260 - val_loss: 0.1798 - val_acc: 0.9252\n",
      "Epoch 122/150\n",
      "Epoch 00122: val_loss did not improve\n",
      " - 12s - loss: 0.1567 - acc: 0.9359 - val_loss: 0.2301 - val_acc: 0.8816\n",
      "Epoch 123/150\n",
      "Epoch 00123: val_loss did not improve\n",
      " - 12s - loss: 0.1891 - acc: 0.9143 - val_loss: 0.2816 - val_acc: 0.8847\n",
      "Epoch 124/150\n",
      "Epoch 00124: val_loss did not improve\n",
      " - 12s - loss: 0.2172 - acc: 0.9036 - val_loss: 0.2029 - val_acc: 0.9159\n",
      "Epoch 125/150\n",
      "Epoch 00125: val_loss did not improve\n",
      " - 12s - loss: 0.1732 - acc: 0.9283 - val_loss: 0.2053 - val_acc: 0.9190\n",
      "Epoch 126/150\n",
      "Epoch 00126: val_loss did not improve\n",
      " - 12s - loss: 0.1646 - acc: 0.9253 - val_loss: 0.2011 - val_acc: 0.9221\n",
      "Epoch 127/150\n",
      "Epoch 00127: val_loss did not improve\n",
      " - 12s - loss: 0.1413 - acc: 0.9436 - val_loss: 0.2078 - val_acc: 0.9065\n",
      "Epoch 128/150\n",
      "Epoch 00128: val_loss did not improve\n",
      " - 12s - loss: 0.1804 - acc: 0.9242 - val_loss: 0.2443 - val_acc: 0.8972\n",
      "Epoch 129/150\n",
      "Epoch 00129: val_loss did not improve\n",
      " - 12s - loss: 0.1994 - acc: 0.9169 - val_loss: 0.1917 - val_acc: 0.9190\n",
      "Epoch 130/150\n",
      "Epoch 00130: val_loss did not improve\n",
      " - 12s - loss: 0.1560 - acc: 0.9382 - val_loss: 0.2010 - val_acc: 0.9221\n",
      "Epoch 131/150\n",
      "Epoch 00131: val_loss did not improve\n",
      " - 12s - loss: 0.1798 - acc: 0.9188 - val_loss: 0.1800 - val_acc: 0.9283\n",
      "Epoch 132/150\n",
      "Epoch 00132: val_loss did not improve\n",
      " - 12s - loss: 0.1626 - acc: 0.9276 - val_loss: 0.2126 - val_acc: 0.9097\n",
      "Epoch 133/150\n",
      "Epoch 00133: val_loss did not improve\n",
      " - 12s - loss: 0.1480 - acc: 0.9367 - val_loss: 0.1809 - val_acc: 0.9221\n",
      "Epoch 134/150\n",
      "Epoch 00134: val_loss did not improve\n",
      " - 12s - loss: 0.1538 - acc: 0.9337 - val_loss: 0.2189 - val_acc: 0.9190\n",
      "Epoch 135/150\n",
      "Epoch 00135: val_loss did not improve\n",
      " - 12s - loss: 0.1672 - acc: 0.9245 - val_loss: 0.2118 - val_acc: 0.9128\n",
      "Epoch 136/150\n",
      "Epoch 00136: val_loss did not improve\n",
      " - 12s - loss: 0.1638 - acc: 0.9337 - val_loss: 0.2060 - val_acc: 0.9128\n",
      "Epoch 137/150\n",
      "Epoch 00137: val_loss did not improve\n",
      " - 12s - loss: 0.1855 - acc: 0.9204 - val_loss: 0.1910 - val_acc: 0.9346\n",
      "Epoch 138/150\n",
      "Epoch 00138: val_loss did not improve\n",
      " - 12s - loss: 0.1973 - acc: 0.9097 - val_loss: 0.1996 - val_acc: 0.9190\n",
      "Epoch 139/150\n",
      "Epoch 00139: val_loss did not improve\n",
      " - 12s - loss: 0.1655 - acc: 0.9237 - val_loss: 0.1885 - val_acc: 0.9221\n",
      "Epoch 140/150\n",
      "Epoch 00140: val_loss did not improve\n",
      " - 12s - loss: 0.1759 - acc: 0.9173 - val_loss: 0.2082 - val_acc: 0.8879\n",
      "Epoch 141/150\n",
      "Epoch 00141: val_loss did not improve\n",
      " - 12s - loss: 0.1507 - acc: 0.9443 - val_loss: 0.1842 - val_acc: 0.9190\n",
      "Epoch 142/150\n",
      "Epoch 00142: val_loss did not improve\n",
      " - 12s - loss: 0.1520 - acc: 0.9413 - val_loss: 0.1966 - val_acc: 0.9190\n",
      "Epoch 143/150\n",
      "Epoch 00143: val_loss did not improve\n",
      " - 12s - loss: 0.1551 - acc: 0.9295 - val_loss: 0.1992 - val_acc: 0.9283\n",
      "Epoch 144/150\n",
      "Epoch 00144: val_loss did not improve\n",
      " - 12s - loss: 0.1669 - acc: 0.9234 - val_loss: 0.1857 - val_acc: 0.9221\n",
      "Epoch 145/150\n",
      "Epoch 00145: val_loss did not improve\n",
      " - 12s - loss: 0.1444 - acc: 0.9413 - val_loss: 0.2109 - val_acc: 0.9097\n",
      "Epoch 146/150\n",
      "Epoch 00146: val_loss did not improve\n",
      " - 12s - loss: 0.1418 - acc: 0.9405 - val_loss: 0.2004 - val_acc: 0.9252\n",
      "Epoch 147/150\n",
      "Epoch 00147: val_loss did not improve\n",
      " - 12s - loss: 0.1348 - acc: 0.9375 - val_loss: 0.2246 - val_acc: 0.9034\n",
      "Epoch 148/150\n",
      "Epoch 00148: val_loss did not improve\n",
      " - 12s - loss: 0.1348 - acc: 0.9352 - val_loss: 0.2349 - val_acc: 0.9065\n",
      "Epoch 149/150\n",
      "Epoch 00149: val_loss did not improve\n",
      " - 12s - loss: 0.1332 - acc: 0.9428 - val_loss: 0.1845 - val_acc: 0.9221\n",
      "Epoch 150/150\n",
      "Epoch 00150: val_loss did not improve\n",
      " - 12s - loss: 0.1495 - acc: 0.9344 - val_loss: 0.2112 - val_acc: 0.9097\n",
      "Satge 2,done!\n",
      "1283/1283 [==============================] - 3s 3ms/step\n",
      "Train loss: 0.138439359104\n",
      "Train accuracy: 0.94855806703\n",
      "321/321 [==============================] - 1s 3ms/step\n",
      "Valid loss: 0.179796458831\n",
      "Valid accuracy: 0.92523364486\n",
      "\n",
      "===================FOLD= 3\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_3 (InputLayer)         (None, 75, 75, 3)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_9 (Conv2D)            (None, 71, 71, 128)       9728      \n",
      "_________________________________________________________________\n",
      "activation_9 (Activation)    (None, 71, 71, 128)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_9 (MaxPooling2 (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "dropout_13 (Dropout)         (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_10 (Conv2D)           (None, 33, 33, 256)       295168    \n",
      "_________________________________________________________________\n",
      "activation_10 (Activation)   (None, 33, 33, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_10 (MaxPooling (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "dropout_14 (Dropout)         (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_11 (Conv2D)           (None, 14, 14, 256)       590080    \n",
      "_________________________________________________________________\n",
      "activation_11 (Activation)   (None, 14, 14, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_11 (MaxPooling (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "dropout_15 (Dropout)         (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_12 (Conv2D)           (None, 5, 5, 128)         295040    \n",
      "_________________________________________________________________\n",
      "activation_12 (Activation)   (None, 5, 5, 128)         0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_12 (MaxPooling (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "dropout_16 (Dropout)         (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten_3 (Flatten)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_7 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dropout_17 (Dropout)         (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_8 (Dense)              (None, 256)               131328    \n",
      "_________________________________________________________________\n",
      "dropout_18 (Dropout)         (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_9 (Dense)              (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 1,584,257\n",
      "Trainable params: 1,584,257\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/150\n",
      "Epoch 00001: val_loss improved from inf to 0.68817, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 14s - loss: 1.1491 - acc: 0.5162 - val_loss: 0.6882 - val_acc: 0.5576\n",
      "Epoch 2/150\n",
      "Epoch 00002: val_loss improved from 0.68817 to 0.67044, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.7307 - acc: 0.5250 - val_loss: 0.6704 - val_acc: 0.5109\n",
      "Epoch 3/150\n",
      "Epoch 00003: val_loss improved from 0.67044 to 0.59148, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.6735 - acc: 0.5902 - val_loss: 0.5915 - val_acc: 0.6667\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 4/150\n",
      "Epoch 00004: val_loss improved from 0.59148 to 0.55713, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.5985 - acc: 0.6210 - val_loss: 0.5571 - val_acc: 0.6978\n",
      "Epoch 5/150\n",
      "Epoch 00005: val_loss improved from 0.55713 to 0.54717, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.5834 - acc: 0.6313 - val_loss: 0.5472 - val_acc: 0.7040\n",
      "Epoch 6/150\n",
      "Epoch 00006: val_loss improved from 0.54717 to 0.53486, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.5695 - acc: 0.6545 - val_loss: 0.5349 - val_acc: 0.7072\n",
      "Epoch 7/150\n",
      "Epoch 00007: val_loss improved from 0.53486 to 0.52415, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.5568 - acc: 0.6634 - val_loss: 0.5242 - val_acc: 0.7196\n",
      "Epoch 8/150\n",
      "Epoch 00008: val_loss improved from 0.52415 to 0.52155, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.5618 - acc: 0.6478 - val_loss: 0.5216 - val_acc: 0.7477\n",
      "Epoch 9/150\n",
      "Epoch 00009: val_loss improved from 0.52155 to 0.50104, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.5279 - acc: 0.6867 - val_loss: 0.5010 - val_acc: 0.7570\n",
      "Epoch 10/150\n",
      "Epoch 00010: val_loss did not improve\n",
      " - 12s - loss: 0.5239 - acc: 0.7099 - val_loss: 0.5095 - val_acc: 0.7477\n",
      "Epoch 11/150\n",
      "Epoch 00011: val_loss improved from 0.50104 to 0.47857, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.4929 - acc: 0.7625 - val_loss: 0.4786 - val_acc: 0.7882\n",
      "Epoch 12/150\n",
      "Epoch 00012: val_loss improved from 0.47857 to 0.44078, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.4481 - acc: 0.7781 - val_loss: 0.4408 - val_acc: 0.8162\n",
      "Epoch 13/150\n",
      "Epoch 00013: val_loss did not improve\n",
      " - 12s - loss: 0.4792 - acc: 0.7625 - val_loss: 0.4661 - val_acc: 0.8037\n",
      "Epoch 14/150\n",
      "Epoch 00014: val_loss improved from 0.44078 to 0.40685, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.4359 - acc: 0.7953 - val_loss: 0.4068 - val_acc: 0.8287\n",
      "Epoch 15/150\n",
      "Epoch 00015: val_loss did not improve\n",
      " - 12s - loss: 0.4131 - acc: 0.8011 - val_loss: 0.4237 - val_acc: 0.8287\n",
      "Epoch 16/150\n",
      "Epoch 00016: val_loss did not improve\n",
      " - 12s - loss: 0.3904 - acc: 0.8105 - val_loss: 0.4633 - val_acc: 0.7695\n",
      "Epoch 17/150\n",
      "Epoch 00017: val_loss improved from 0.40685 to 0.39952, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.3783 - acc: 0.8330 - val_loss: 0.3995 - val_acc: 0.8318\n",
      "Epoch 18/150\n",
      "Epoch 00018: val_loss improved from 0.39952 to 0.39796, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.3722 - acc: 0.8296 - val_loss: 0.3980 - val_acc: 0.8442\n",
      "Epoch 19/150\n",
      "Epoch 00019: val_loss did not improve\n",
      " - 12s - loss: 0.3724 - acc: 0.8338 - val_loss: 0.4011 - val_acc: 0.8100\n",
      "Epoch 20/150\n",
      "Epoch 00020: val_loss improved from 0.39796 to 0.36801, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.3512 - acc: 0.8429 - val_loss: 0.3680 - val_acc: 0.8380\n",
      "Epoch 21/150\n",
      "Epoch 00021: val_loss did not improve\n",
      " - 12s - loss: 0.3667 - acc: 0.8311 - val_loss: 0.3761 - val_acc: 0.8474\n",
      "Epoch 22/150\n",
      "Epoch 00022: val_loss did not improve\n",
      " - 12s - loss: 0.3654 - acc: 0.8349 - val_loss: 0.3870 - val_acc: 0.8069\n",
      "Epoch 23/150\n",
      "Epoch 00023: val_loss improved from 0.36801 to 0.36749, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.3437 - acc: 0.8536 - val_loss: 0.3675 - val_acc: 0.8131\n",
      "Epoch 24/150\n",
      "Epoch 00024: val_loss improved from 0.36749 to 0.34548, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.3405 - acc: 0.8410 - val_loss: 0.3455 - val_acc: 0.8536\n",
      "Epoch 25/150\n",
      "Epoch 00025: val_loss improved from 0.34548 to 0.31503, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.3258 - acc: 0.8559 - val_loss: 0.3150 - val_acc: 0.8692\n",
      "Epoch 26/150\n",
      "Epoch 00026: val_loss improved from 0.31503 to 0.30904, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.3134 - acc: 0.8635 - val_loss: 0.3090 - val_acc: 0.8536\n",
      "Epoch 27/150\n",
      "Epoch 00027: val_loss did not improve\n",
      " - 12s - loss: 0.3161 - acc: 0.8449 - val_loss: 0.3887 - val_acc: 0.7788\n",
      "Epoch 28/150\n",
      "Epoch 00028: val_loss did not improve\n",
      " - 12s - loss: 0.3306 - acc: 0.8342 - val_loss: 0.3162 - val_acc: 0.8567\n",
      "Epoch 29/150\n",
      "Epoch 00029: val_loss improved from 0.30904 to 0.30458, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.3011 - acc: 0.8536 - val_loss: 0.3046 - val_acc: 0.8723\n",
      "Epoch 30/150\n",
      "Epoch 00030: val_loss did not improve\n",
      " - 12s - loss: 0.2887 - acc: 0.8627 - val_loss: 0.3211 - val_acc: 0.8598\n",
      "Epoch 31/150\n",
      "Epoch 00031: val_loss did not improve\n",
      " - 12s - loss: 0.2897 - acc: 0.8711 - val_loss: 0.3120 - val_acc: 0.8474\n",
      "Epoch 32/150\n",
      "Epoch 00032: val_loss improved from 0.30458 to 0.28157, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.2713 - acc: 0.8795 - val_loss: 0.2816 - val_acc: 0.8879\n",
      "Epoch 33/150\n",
      "Epoch 00033: val_loss did not improve\n",
      " - 12s - loss: 0.2771 - acc: 0.8731 - val_loss: 0.3123 - val_acc: 0.8692\n",
      "Epoch 34/150\n",
      "Epoch 00034: val_loss improved from 0.28157 to 0.28145, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.2742 - acc: 0.8810 - val_loss: 0.2815 - val_acc: 0.8754\n",
      "Epoch 35/150\n",
      "Epoch 00035: val_loss did not improve\n",
      " - 12s - loss: 0.2851 - acc: 0.8590 - val_loss: 0.3056 - val_acc: 0.8660\n",
      "Epoch 36/150\n",
      "Epoch 00036: val_loss did not improve\n",
      " - 12s - loss: 0.2628 - acc: 0.8795 - val_loss: 0.2853 - val_acc: 0.8723\n",
      "Epoch 37/150\n",
      "Epoch 00037: val_loss did not improve\n",
      " - 12s - loss: 0.2993 - acc: 0.8559 - val_loss: 0.2957 - val_acc: 0.8660\n",
      "Epoch 38/150\n",
      "Epoch 00038: val_loss did not improve\n",
      " - 12s - loss: 0.2445 - acc: 0.8848 - val_loss: 0.2863 - val_acc: 0.8816\n",
      "Epoch 39/150\n",
      "Epoch 00039: val_loss improved from 0.28145 to 0.27618, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.2579 - acc: 0.8772 - val_loss: 0.2762 - val_acc: 0.8941\n",
      "Epoch 40/150\n",
      "Epoch 00040: val_loss improved from 0.27618 to 0.26448, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv3.hdf5\n",
      " - 12s - loss: 0.2474 - acc: 0.8856 - val_loss: 0.2645 - val_acc: 0.8847\n",
      "Epoch 41/150\n",
      "Epoch 00041: val_loss did not improve\n",
      " - 12s - loss: 0.2491 - acc: 0.8902 - val_loss: 0.2851 - val_acc: 0.8629\n",
      "Epoch 42/150\n",
      "Epoch 00042: val_loss did not improve\n",
      " - 12s - loss: 0.2524 - acc: 0.8807 - val_loss: 0.2954 - val_acc: 0.8598\n",
      "Epoch 43/150\n",
      "Epoch 00043: val_loss did not improve\n",
      " - 12s - loss: 0.2486 - acc: 0.8826 - val_loss: 0.2839 - val_acc: 0.8754\n",
      "Epoch 44/150\n",
      "Epoch 00044: val_loss did not improve\n",
      " - 12s - loss: 0.2582 - acc: 0.8887 - val_loss: 0.2746 - val_acc: 0.8723\n",
      "Epoch 45/150\n",
      "Epoch 00045: val_loss did not improve\n",
      " - 12s - loss: 0.2901 - acc: 0.8643 - val_loss: 0.4099 - val_acc: 0.7601\n",
      "Epoch 46/150\n",
      "Epoch 00046: val_loss did not improve\n",
      " - 12s - loss: 0.2746 - acc: 0.8734 - val_loss: 0.2707 - val_acc: 0.8847\n",
      "Epoch 47/150\n",
      "Epoch 00047: val_loss did not improve\n",
      " - 12s - loss: 0.2744 - acc: 0.8781 - val_loss: 0.3413 - val_acc: 0.8754\n",
      "Epoch 48/150\n",
      "Epoch 00048: val_loss did not improve\n",
      " - 12s - loss: 0.2710 - acc: 0.8765 - val_loss: 0.2920 - val_acc: 0.8847\n",
      "Epoch 49/150\n",
      "Epoch 00049: val_loss did not improve\n",
      " - 12s - loss: 0.2335 - acc: 0.8887 - val_loss: 0.2749 - val_acc: 0.8879\n",
      "Epoch 50/150\n",
      "Epoch 00050: val_loss did not improve\n",
      " - 12s - loss: 0.2410 - acc: 0.8899 - val_loss: 0.2960 - val_acc: 0.8629\n",
      "Epoch 51/150\n",
      "Epoch 00051: val_loss did not improve\n",
      " - 12s - loss: 0.2516 - acc: 0.8894 - val_loss: 0.2705 - val_acc: 0.8816\n",
      "Epoch 52/150\n",
      "Epoch 00052: val_loss did not improve\n",
      " - 12s - loss: 0.2417 - acc: 0.8887 - val_loss: 0.3203 - val_acc: 0.8692\n",
      "Epoch 53/150\n",
      "Epoch 00053: val_loss did not improve\n",
      " - 12s - loss: 0.2318 - acc: 0.8917 - val_loss: 0.2752 - val_acc: 0.8785\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 54/150\n",
      "Epoch 00054: val_loss did not improve\n",
      " - 12s - loss: 0.2307 - acc: 0.8929 - val_loss: 0.2971 - val_acc: 0.8598\n",
      "Epoch 55/150\n",
      "Epoch 00055: val_loss did not improve\n",
      " - 12s - loss: 0.2338 - acc: 0.8810 - val_loss: 0.2870 - val_acc: 0.8816\n",
      "Epoch 56/150\n",
      "Epoch 00056: val_loss did not improve\n",
      " - 12s - loss: 0.2426 - acc: 0.8860 - val_loss: 0.2734 - val_acc: 0.8816\n",
      "Epoch 57/150\n",
      "Epoch 00057: val_loss did not improve\n",
      " - 12s - loss: 0.2417 - acc: 0.8879 - val_loss: 0.2785 - val_acc: 0.8723\n",
      "Epoch 58/150\n",
      "Epoch 00058: val_loss did not improve\n",
      " - 12s - loss: 0.2268 - acc: 0.8978 - val_loss: 0.2714 - val_acc: 0.8723\n",
      "Epoch 59/150\n",
      "Epoch 00059: val_loss did not improve\n",
      " - 12s - loss: 0.2241 - acc: 0.9036 - val_loss: 0.3285 - val_acc: 0.8785\n",
      "Epoch 60/150\n",
      "Epoch 00060: val_loss did not improve\n",
      " - 12s - loss: 0.2235 - acc: 0.8993 - val_loss: 0.2855 - val_acc: 0.8754\n",
      "Epoch 61/150\n",
      "Epoch 00061: val_loss did not improve\n",
      " - 12s - loss: 0.2078 - acc: 0.9070 - val_loss: 0.2843 - val_acc: 0.8910\n",
      "Epoch 62/150\n",
      "Epoch 00062: val_loss did not improve\n",
      " - 12s - loss: 0.2298 - acc: 0.9039 - val_loss: 0.3114 - val_acc: 0.8505\n",
      "Epoch 63/150\n",
      "Epoch 00063: val_loss did not improve\n",
      " - 12s - loss: 0.2075 - acc: 0.8978 - val_loss: 0.2930 - val_acc: 0.8816\n",
      "Epoch 64/150\n",
      "Epoch 00064: val_loss did not improve\n",
      " - 12s - loss: 0.2063 - acc: 0.9146 - val_loss: 0.2758 - val_acc: 0.8972\n",
      "Epoch 65/150\n",
      "Epoch 00065: val_loss did not improve\n",
      " - 12s - loss: 0.1961 - acc: 0.9154 - val_loss: 0.2987 - val_acc: 0.8847\n",
      "Epoch 66/150\n",
      "Epoch 00066: val_loss did not improve\n",
      " - 12s - loss: 0.2063 - acc: 0.9100 - val_loss: 0.3001 - val_acc: 0.8816\n",
      "Epoch 67/150\n",
      "Epoch 00067: val_loss did not improve\n",
      " - 12s - loss: 0.2019 - acc: 0.9077 - val_loss: 0.2817 - val_acc: 0.8941\n",
      "Epoch 68/150\n",
      "Epoch 00068: val_loss did not improve\n",
      " - 12s - loss: 0.2217 - acc: 0.9092 - val_loss: 0.2817 - val_acc: 0.8785\n",
      "Epoch 69/150\n",
      "Epoch 00069: val_loss did not improve\n",
      " - 12s - loss: 0.2087 - acc: 0.9016 - val_loss: 0.2880 - val_acc: 0.8816\n",
      "Epoch 70/150\n",
      "Epoch 00070: val_loss did not improve\n",
      " - 12s - loss: 0.1925 - acc: 0.9184 - val_loss: 0.2864 - val_acc: 0.8972\n",
      "Satge 2,done!\n",
      "1283/1283 [==============================] - 3s 3ms/step\n",
      "Train loss: 0.216248116987\n",
      "Train accuracy: 0.90958690569\n",
      "321/321 [==============================] - 1s 3ms/step\n",
      "Valid loss: 0.264476769122\n",
      "Valid accuracy: 0.884735202492\n",
      "\n",
      "===================FOLD= 4\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_4 (InputLayer)         (None, 75, 75, 3)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_13 (Conv2D)           (None, 71, 71, 128)       9728      \n",
      "_________________________________________________________________\n",
      "activation_13 (Activation)   (None, 71, 71, 128)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_13 (MaxPooling (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "dropout_19 (Dropout)         (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_14 (Conv2D)           (None, 33, 33, 256)       295168    \n",
      "_________________________________________________________________\n",
      "activation_14 (Activation)   (None, 33, 33, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_14 (MaxPooling (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "dropout_20 (Dropout)         (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_15 (Conv2D)           (None, 14, 14, 256)       590080    \n",
      "_________________________________________________________________\n",
      "activation_15 (Activation)   (None, 14, 14, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_15 (MaxPooling (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "dropout_21 (Dropout)         (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_16 (Conv2D)           (None, 5, 5, 128)         295040    \n",
      "_________________________________________________________________\n",
      "activation_16 (Activation)   (None, 5, 5, 128)         0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_16 (MaxPooling (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "dropout_22 (Dropout)         (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten_4 (Flatten)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_10 (Dense)             (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dropout_23 (Dropout)         (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_11 (Dense)             (None, 256)               131328    \n",
      "_________________________________________________________________\n",
      "dropout_24 (Dropout)         (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_12 (Dense)             (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 1,584,257\n",
      "Trainable params: 1,584,257\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/150\n",
      "Epoch 00001: val_loss improved from inf to 0.68111, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 15s - loss: 1.3876 - acc: 0.5060 - val_loss: 0.6811 - val_acc: 0.5437\n",
      "Epoch 2/150\n",
      "Epoch 00002: val_loss did not improve\n",
      " - 12s - loss: 0.7186 - acc: 0.5250 - val_loss: 0.6829 - val_acc: 0.5469\n",
      "Epoch 3/150\n",
      "Epoch 00003: val_loss improved from 0.68111 to 0.65566, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.6965 - acc: 0.5428 - val_loss: 0.6557 - val_acc: 0.5531\n",
      "Epoch 4/150\n",
      "Epoch 00004: val_loss improved from 0.65566 to 0.59820, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.6355 - acc: 0.5862 - val_loss: 0.5982 - val_acc: 0.5312\n",
      "Epoch 5/150\n",
      "Epoch 00005: val_loss improved from 0.59820 to 0.57260, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.5785 - acc: 0.6458 - val_loss: 0.5726 - val_acc: 0.6469\n",
      "Epoch 6/150\n",
      "Epoch 00006: val_loss did not improve\n",
      " - 12s - loss: 0.5627 - acc: 0.6601 - val_loss: 0.5742 - val_acc: 0.6594\n",
      "Epoch 7/150\n",
      "Epoch 00007: val_loss improved from 0.57260 to 0.56445, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.5541 - acc: 0.6798 - val_loss: 0.5645 - val_acc: 0.6719\n",
      "Epoch 8/150\n",
      "Epoch 00008: val_loss improved from 0.56445 to 0.54546, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.5209 - acc: 0.7065 - val_loss: 0.5455 - val_acc: 0.7156\n",
      "Epoch 9/150\n",
      "Epoch 00009: val_loss improved from 0.54546 to 0.54170, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.5290 - acc: 0.7028 - val_loss: 0.5417 - val_acc: 0.7344\n",
      "Epoch 10/150\n",
      "Epoch 00010: val_loss improved from 0.54170 to 0.50994, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.4852 - acc: 0.7499 - val_loss: 0.5099 - val_acc: 0.7656\n",
      "Epoch 11/150\n",
      "Epoch 00011: val_loss improved from 0.50994 to 0.50603, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.4659 - acc: 0.7781 - val_loss: 0.5060 - val_acc: 0.7906\n",
      "Epoch 12/150\n",
      "Epoch 00012: val_loss improved from 0.50603 to 0.50181, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.4633 - acc: 0.7669 - val_loss: 0.5018 - val_acc: 0.7844\n",
      "Epoch 13/150\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 00013: val_loss improved from 0.50181 to 0.46166, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.4320 - acc: 0.7987 - val_loss: 0.4617 - val_acc: 0.8000\n",
      "Epoch 14/150\n",
      "Epoch 00014: val_loss improved from 0.46166 to 0.44577, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.4472 - acc: 0.7761 - val_loss: 0.4458 - val_acc: 0.8187\n",
      "Epoch 15/150\n",
      "Epoch 00015: val_loss did not improve\n",
      " - 12s - loss: 0.4109 - acc: 0.8078 - val_loss: 0.4680 - val_acc: 0.7719\n",
      "Epoch 16/150\n",
      "Epoch 00016: val_loss did not improve\n",
      " - 12s - loss: 0.4292 - acc: 0.7956 - val_loss: 0.4492 - val_acc: 0.7906\n",
      "Epoch 17/150\n",
      "Epoch 00017: val_loss did not improve\n",
      " - 12s - loss: 0.3886 - acc: 0.8216 - val_loss: 0.4605 - val_acc: 0.7875\n",
      "Epoch 18/150\n",
      "Epoch 00018: val_loss improved from 0.44577 to 0.38677, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.3732 - acc: 0.8179 - val_loss: 0.3868 - val_acc: 0.8406\n",
      "Epoch 19/150\n",
      "Epoch 00019: val_loss did not improve\n",
      " - 12s - loss: 0.3631 - acc: 0.8354 - val_loss: 0.4598 - val_acc: 0.8031\n",
      "Epoch 20/150\n",
      "Epoch 00020: val_loss improved from 0.38677 to 0.37618, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.3489 - acc: 0.8476 - val_loss: 0.3762 - val_acc: 0.8250\n",
      "Epoch 21/150\n",
      "Epoch 00021: val_loss did not improve\n",
      " - 12s - loss: 0.3539 - acc: 0.8406 - val_loss: 0.3917 - val_acc: 0.8281\n",
      "Epoch 22/150\n",
      "Epoch 00022: val_loss did not improve\n",
      " - 12s - loss: 0.3430 - acc: 0.8453 - val_loss: 0.3933 - val_acc: 0.8281\n",
      "Epoch 23/150\n",
      "Epoch 00023: val_loss improved from 0.37618 to 0.33081, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.3205 - acc: 0.8476 - val_loss: 0.3308 - val_acc: 0.8656\n",
      "Epoch 24/150\n",
      "Epoch 00024: val_loss improved from 0.33081 to 0.30562, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.3353 - acc: 0.8407 - val_loss: 0.3056 - val_acc: 0.8562\n",
      "Epoch 25/150\n",
      "Epoch 00025: val_loss improved from 0.30562 to 0.30386, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.3229 - acc: 0.8498 - val_loss: 0.3039 - val_acc: 0.8656\n",
      "Epoch 26/150\n",
      "Epoch 00026: val_loss did not improve\n",
      " - 12s - loss: 0.3396 - acc: 0.8545 - val_loss: 0.3462 - val_acc: 0.8156\n",
      "Epoch 27/150\n",
      "Epoch 00027: val_loss did not improve\n",
      " - 12s - loss: 0.3051 - acc: 0.8643 - val_loss: 0.3148 - val_acc: 0.8438\n",
      "Epoch 28/150\n",
      "Epoch 00028: val_loss did not improve\n",
      " - 12s - loss: 0.3178 - acc: 0.8516 - val_loss: 0.3496 - val_acc: 0.8469\n",
      "Epoch 29/150\n",
      "Epoch 00029: val_loss did not improve\n",
      " - 12s - loss: 0.2822 - acc: 0.8826 - val_loss: 0.3905 - val_acc: 0.8594\n",
      "Epoch 30/150\n",
      "Epoch 00030: val_loss did not improve\n",
      " - 12s - loss: 0.3614 - acc: 0.8362 - val_loss: 0.3169 - val_acc: 0.8500\n",
      "Epoch 31/150\n",
      "Epoch 00031: val_loss did not improve\n",
      " - 12s - loss: 0.2981 - acc: 0.8612 - val_loss: 0.3046 - val_acc: 0.8625\n",
      "Epoch 32/150\n",
      "Epoch 00032: val_loss did not improve\n",
      " - 12s - loss: 0.2953 - acc: 0.8537 - val_loss: 0.3098 - val_acc: 0.8688\n",
      "Epoch 33/150\n",
      "Epoch 00033: val_loss improved from 0.30386 to 0.29035, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.3020 - acc: 0.8446 - val_loss: 0.2903 - val_acc: 0.8469\n",
      "Epoch 34/150\n",
      "Epoch 00034: val_loss improved from 0.29035 to 0.28502, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.2763 - acc: 0.8721 - val_loss: 0.2850 - val_acc: 0.8594\n",
      "Epoch 35/150\n",
      "Epoch 00035: val_loss did not improve\n",
      " - 12s - loss: 0.2687 - acc: 0.8795 - val_loss: 0.3560 - val_acc: 0.8156\n",
      "Epoch 36/150\n",
      "Epoch 00036: val_loss did not improve\n",
      " - 12s - loss: 0.2864 - acc: 0.8667 - val_loss: 0.2987 - val_acc: 0.8688\n",
      "Epoch 37/150\n",
      "Epoch 00037: val_loss improved from 0.28502 to 0.27233, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.2754 - acc: 0.8804 - val_loss: 0.2723 - val_acc: 0.8594\n",
      "Epoch 38/150\n",
      "Epoch 00038: val_loss did not improve\n",
      " - 12s - loss: 0.2419 - acc: 0.8948 - val_loss: 0.2753 - val_acc: 0.8781\n",
      "Epoch 39/150\n",
      "Epoch 00039: val_loss did not improve\n",
      " - 12s - loss: 0.2703 - acc: 0.8812 - val_loss: 0.2921 - val_acc: 0.8750\n",
      "Epoch 40/150\n",
      "Epoch 00040: val_loss did not improve\n",
      " - 12s - loss: 0.2921 - acc: 0.8554 - val_loss: 0.3432 - val_acc: 0.8313\n",
      "Epoch 41/150\n",
      "Epoch 00041: val_loss did not improve\n",
      " - 12s - loss: 0.2771 - acc: 0.8597 - val_loss: 0.3284 - val_acc: 0.8250\n",
      "Epoch 42/150\n",
      "Epoch 00042: val_loss improved from 0.27233 to 0.27081, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.2695 - acc: 0.8766 - val_loss: 0.2708 - val_acc: 0.8656\n",
      "Epoch 43/150\n",
      "Epoch 00043: val_loss did not improve\n",
      " - 12s - loss: 0.2422 - acc: 0.8879 - val_loss: 0.2784 - val_acc: 0.8781\n",
      "Epoch 44/150\n",
      "Epoch 00044: val_loss did not improve\n",
      " - 12s - loss: 0.2683 - acc: 0.8871 - val_loss: 0.3385 - val_acc: 0.8125\n",
      "Epoch 45/150\n",
      "Epoch 00045: val_loss improved from 0.27081 to 0.26223, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.2482 - acc: 0.8818 - val_loss: 0.2622 - val_acc: 0.8719\n",
      "Epoch 46/150\n",
      "Epoch 00046: val_loss did not improve\n",
      " - 12s - loss: 0.2578 - acc: 0.8758 - val_loss: 0.2873 - val_acc: 0.8844\n",
      "Epoch 47/150\n",
      "Epoch 00047: val_loss did not improve\n",
      " - 12s - loss: 0.2457 - acc: 0.8871 - val_loss: 0.2715 - val_acc: 0.8875\n",
      "Epoch 48/150\n",
      "Epoch 00048: val_loss did not improve\n",
      " - 12s - loss: 0.3190 - acc: 0.8448 - val_loss: 0.2627 - val_acc: 0.8906\n",
      "Epoch 49/150\n",
      "Epoch 00049: val_loss did not improve\n",
      " - 12s - loss: 0.2493 - acc: 0.8902 - val_loss: 0.2682 - val_acc: 0.8781\n",
      "Epoch 50/150\n",
      "Epoch 00050: val_loss improved from 0.26223 to 0.25504, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.2492 - acc: 0.8918 - val_loss: 0.2550 - val_acc: 0.8844\n",
      "Epoch 51/150\n",
      "Epoch 00051: val_loss improved from 0.25504 to 0.25106, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.2437 - acc: 0.8940 - val_loss: 0.2511 - val_acc: 0.8906\n",
      "Epoch 52/150\n",
      "Epoch 00052: val_loss improved from 0.25106 to 0.23365, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.2382 - acc: 0.8871 - val_loss: 0.2337 - val_acc: 0.8844\n",
      "Epoch 53/150\n",
      "Epoch 00053: val_loss did not improve\n",
      " - 12s - loss: 0.2858 - acc: 0.8674 - val_loss: 0.2607 - val_acc: 0.8625\n",
      "Epoch 54/150\n",
      "Epoch 00054: val_loss did not improve\n",
      " - 12s - loss: 0.2325 - acc: 0.8925 - val_loss: 0.2505 - val_acc: 0.8781\n",
      "Epoch 55/150\n",
      "Epoch 00055: val_loss did not improve\n",
      " - 12s - loss: 0.2367 - acc: 0.8972 - val_loss: 0.2981 - val_acc: 0.8719\n",
      "Epoch 56/150\n",
      "Epoch 00056: val_loss did not improve\n",
      " - 12s - loss: 0.2422 - acc: 0.8887 - val_loss: 0.2403 - val_acc: 0.8969\n",
      "Epoch 57/150\n",
      "Epoch 00057: val_loss did not improve\n",
      " - 12s - loss: 0.2423 - acc: 0.8842 - val_loss: 0.2513 - val_acc: 0.8812\n",
      "Epoch 58/150\n",
      "Epoch 00058: val_loss did not improve\n",
      " - 12s - loss: 0.2446 - acc: 0.8827 - val_loss: 0.2913 - val_acc: 0.8719\n",
      "Epoch 59/150\n",
      "Epoch 00059: val_loss did not improve\n",
      " - 12s - loss: 0.2599 - acc: 0.8971 - val_loss: 0.2654 - val_acc: 0.8781\n",
      "Epoch 60/150\n",
      "Epoch 00060: val_loss did not improve\n",
      " - 12s - loss: 0.2194 - acc: 0.9047 - val_loss: 0.2637 - val_acc: 0.8875\n",
      "Epoch 61/150\n",
      "Epoch 00061: val_loss did not improve\n",
      " - 12s - loss: 0.2577 - acc: 0.8728 - val_loss: 0.2604 - val_acc: 0.8781\n",
      "Epoch 62/150\n",
      "Epoch 00062: val_loss did not improve\n",
      " - 12s - loss: 0.2210 - acc: 0.8903 - val_loss: 0.2418 - val_acc: 0.8875\n",
      "Epoch 63/150\n",
      "Epoch 00063: val_loss did not improve\n",
      " - 12s - loss: 0.2303 - acc: 0.8917 - val_loss: 0.2361 - val_acc: 0.8938\n",
      "Epoch 64/150\n",
      "Epoch 00064: val_loss did not improve\n",
      " - 12s - loss: 0.2173 - acc: 0.9115 - val_loss: 0.2436 - val_acc: 0.8906\n",
      "Epoch 65/150\n",
      "Epoch 00065: val_loss did not improve\n",
      " - 12s - loss: 0.2319 - acc: 0.8927 - val_loss: 0.2355 - val_acc: 0.8875\n",
      "Epoch 66/150\n",
      "Epoch 00066: val_loss improved from 0.23365 to 0.22218, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.2331 - acc: 0.8955 - val_loss: 0.2222 - val_acc: 0.9031\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 67/150\n",
      "Epoch 00067: val_loss did not improve\n",
      " - 12s - loss: 0.2192 - acc: 0.9062 - val_loss: 0.2435 - val_acc: 0.8906\n",
      "Epoch 68/150\n",
      "Epoch 00068: val_loss did not improve\n",
      " - 12s - loss: 0.2188 - acc: 0.9085 - val_loss: 0.2376 - val_acc: 0.8875\n",
      "Epoch 69/150\n",
      "Epoch 00069: val_loss did not improve\n",
      " - 12s - loss: 0.2213 - acc: 0.8986 - val_loss: 0.2241 - val_acc: 0.8906\n",
      "Epoch 70/150\n",
      "Epoch 00070: val_loss did not improve\n",
      " - 12s - loss: 0.2430 - acc: 0.8775 - val_loss: 0.2491 - val_acc: 0.9031\n",
      "Epoch 71/150\n",
      "Epoch 00071: val_loss did not improve\n",
      " - 12s - loss: 0.2239 - acc: 0.9054 - val_loss: 0.2396 - val_acc: 0.8969\n",
      "Epoch 72/150\n",
      "Epoch 00072: val_loss did not improve\n",
      " - 12s - loss: 0.2044 - acc: 0.9039 - val_loss: 0.2525 - val_acc: 0.8750\n",
      "Epoch 73/150\n",
      "Epoch 00073: val_loss did not improve\n",
      " - 12s - loss: 0.2215 - acc: 0.8963 - val_loss: 0.2528 - val_acc: 0.8906\n",
      "Epoch 74/150\n",
      "Epoch 00074: val_loss did not improve\n",
      " - 12s - loss: 0.2166 - acc: 0.8993 - val_loss: 0.2453 - val_acc: 0.8906\n",
      "Epoch 75/150\n",
      "Epoch 00075: val_loss did not improve\n",
      " - 13s - loss: 0.2239 - acc: 0.8925 - val_loss: 0.2304 - val_acc: 0.8969\n",
      "Epoch 76/150\n",
      "Epoch 00076: val_loss did not improve\n",
      " - 12s - loss: 0.2194 - acc: 0.9032 - val_loss: 0.2575 - val_acc: 0.8875\n",
      "Epoch 77/150\n",
      "Epoch 00077: val_loss did not improve\n",
      " - 12s - loss: 0.1995 - acc: 0.9161 - val_loss: 0.2348 - val_acc: 0.8906\n",
      "Epoch 78/150\n",
      "Epoch 00078: val_loss did not improve\n",
      " - 12s - loss: 0.2199 - acc: 0.9001 - val_loss: 0.2478 - val_acc: 0.8844\n",
      "Epoch 79/150\n",
      "Epoch 00079: val_loss did not improve\n",
      " - 12s - loss: 0.2082 - acc: 0.9009 - val_loss: 0.2241 - val_acc: 0.8938\n",
      "Epoch 80/150\n",
      "Epoch 00080: val_loss did not improve\n",
      " - 12s - loss: 0.2134 - acc: 0.9184 - val_loss: 0.2302 - val_acc: 0.8969\n",
      "Epoch 81/150\n",
      "Epoch 00081: val_loss did not improve\n",
      " - 12s - loss: 0.1973 - acc: 0.9161 - val_loss: 0.2968 - val_acc: 0.8750\n",
      "Epoch 82/150\n",
      "Epoch 00082: val_loss did not improve\n",
      " - 12s - loss: 0.2049 - acc: 0.9100 - val_loss: 0.2366 - val_acc: 0.8906\n",
      "Epoch 83/150\n",
      "Epoch 00083: val_loss did not improve\n",
      " - 12s - loss: 0.2147 - acc: 0.9070 - val_loss: 0.2360 - val_acc: 0.8938\n",
      "Epoch 84/150\n",
      "Epoch 00084: val_loss did not improve\n",
      " - 12s - loss: 0.2137 - acc: 0.9070 - val_loss: 0.2416 - val_acc: 0.9031\n",
      "Epoch 85/150\n",
      "Epoch 00085: val_loss improved from 0.22218 to 0.22108, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.2033 - acc: 0.9184 - val_loss: 0.2211 - val_acc: 0.8938\n",
      "Epoch 86/150\n",
      "Epoch 00086: val_loss did not improve\n",
      " - 12s - loss: 0.1957 - acc: 0.9155 - val_loss: 0.2275 - val_acc: 0.8938\n",
      "Epoch 87/150\n",
      "Epoch 00087: val_loss did not improve\n",
      " - 12s - loss: 0.2153 - acc: 0.9100 - val_loss: 0.2314 - val_acc: 0.9000\n",
      "Epoch 88/150\n",
      "Epoch 00088: val_loss did not improve\n",
      " - 12s - loss: 0.2008 - acc: 0.9154 - val_loss: 0.2232 - val_acc: 0.9031\n",
      "Epoch 89/150\n",
      "Epoch 00089: val_loss did not improve\n",
      " - 12s - loss: 0.1880 - acc: 0.9222 - val_loss: 0.2568 - val_acc: 0.8844\n",
      "Epoch 90/150\n",
      "Epoch 00090: val_loss improved from 0.22108 to 0.20475, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.2056 - acc: 0.9207 - val_loss: 0.2047 - val_acc: 0.9156\n",
      "Epoch 91/150\n",
      "Epoch 00091: val_loss did not improve\n",
      " - 12s - loss: 0.1966 - acc: 0.9146 - val_loss: 0.2287 - val_acc: 0.9031\n",
      "Epoch 92/150\n",
      "Epoch 00092: val_loss did not improve\n",
      " - 12s - loss: 0.1929 - acc: 0.9161 - val_loss: 0.2356 - val_acc: 0.9000\n",
      "Epoch 93/150\n",
      "Epoch 00093: val_loss did not improve\n",
      " - 12s - loss: 0.2091 - acc: 0.8964 - val_loss: 0.2463 - val_acc: 0.8969\n",
      "Epoch 94/150\n",
      "Epoch 00094: val_loss did not improve\n",
      " - 12s - loss: 0.1982 - acc: 0.9117 - val_loss: 0.2140 - val_acc: 0.8969\n",
      "Epoch 95/150\n",
      "Epoch 00095: val_loss did not improve\n",
      " - 12s - loss: 0.1907 - acc: 0.9215 - val_loss: 0.2239 - val_acc: 0.8938\n",
      "Epoch 96/150\n",
      "Epoch 00096: val_loss did not improve\n",
      " - 12s - loss: 0.1793 - acc: 0.9199 - val_loss: 0.2111 - val_acc: 0.8906\n",
      "Epoch 97/150\n",
      "Epoch 00097: val_loss improved from 0.20475 to 0.20382, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.2006 - acc: 0.9154 - val_loss: 0.2038 - val_acc: 0.9031\n",
      "Epoch 98/150\n",
      "Epoch 00098: val_loss did not improve\n",
      " - 12s - loss: 0.1844 - acc: 0.9154 - val_loss: 0.2256 - val_acc: 0.9094\n",
      "Epoch 99/150\n",
      "Epoch 00099: val_loss did not improve\n",
      " - 12s - loss: 0.2010 - acc: 0.9109 - val_loss: 0.2591 - val_acc: 0.8812\n",
      "Epoch 100/150\n",
      "Epoch 00100: val_loss did not improve\n",
      " - 12s - loss: 0.2228 - acc: 0.9025 - val_loss: 0.2591 - val_acc: 0.8688\n",
      "Epoch 101/150\n",
      "Epoch 00101: val_loss did not improve\n",
      " - 12s - loss: 0.2066 - acc: 0.9140 - val_loss: 0.2353 - val_acc: 0.8938\n",
      "Epoch 102/150\n",
      "Epoch 00102: val_loss did not improve\n",
      " - 12s - loss: 0.1900 - acc: 0.9169 - val_loss: 0.2243 - val_acc: 0.9000\n",
      "Epoch 103/150\n",
      "Epoch 00103: val_loss did not improve\n",
      " - 12s - loss: 0.1977 - acc: 0.9077 - val_loss: 0.2255 - val_acc: 0.8969\n",
      "Epoch 104/150\n",
      "Epoch 00104: val_loss did not improve\n",
      " - 12s - loss: 0.2065 - acc: 0.9147 - val_loss: 0.2440 - val_acc: 0.8781\n",
      "Epoch 105/150\n",
      "Epoch 00105: val_loss did not improve\n",
      " - 12s - loss: 0.1886 - acc: 0.9207 - val_loss: 0.2411 - val_acc: 0.8969\n",
      "Epoch 106/150\n",
      "Epoch 00106: val_loss did not improve\n",
      " - 12s - loss: 0.1971 - acc: 0.9115 - val_loss: 0.2280 - val_acc: 0.9000\n",
      "Epoch 107/150\n",
      "Epoch 00107: val_loss did not improve\n",
      " - 12s - loss: 0.1896 - acc: 0.9161 - val_loss: 0.2205 - val_acc: 0.9031\n",
      "Epoch 108/150\n",
      "Epoch 00108: val_loss did not improve\n",
      " - 12s - loss: 0.1719 - acc: 0.9230 - val_loss: 0.2268 - val_acc: 0.8875\n",
      "Epoch 109/150\n",
      "Epoch 00109: val_loss did not improve\n",
      " - 12s - loss: 0.1881 - acc: 0.9215 - val_loss: 0.2597 - val_acc: 0.8750\n",
      "Epoch 110/150\n",
      "Epoch 00110: val_loss did not improve\n",
      " - 12s - loss: 0.2337 - acc: 0.8934 - val_loss: 0.2291 - val_acc: 0.8906\n",
      "Epoch 111/150\n",
      "Epoch 00111: val_loss did not improve\n",
      " - 12s - loss: 0.1936 - acc: 0.9184 - val_loss: 0.2178 - val_acc: 0.9000\n",
      "Epoch 112/150\n",
      "Epoch 00112: val_loss did not improve\n",
      " - 12s - loss: 0.1847 - acc: 0.9268 - val_loss: 0.2285 - val_acc: 0.8938\n",
      "Epoch 113/150\n",
      "Epoch 00113: val_loss did not improve\n",
      " - 12s - loss: 0.1758 - acc: 0.9245 - val_loss: 0.2260 - val_acc: 0.9094\n",
      "Epoch 114/150\n",
      "Epoch 00114: val_loss improved from 0.20382 to 0.19719, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv4.hdf5\n",
      " - 12s - loss: 0.1713 - acc: 0.9215 - val_loss: 0.1972 - val_acc: 0.9156\n",
      "Epoch 115/150\n",
      "Epoch 00115: val_loss did not improve\n",
      " - 12s - loss: 0.1701 - acc: 0.9260 - val_loss: 0.2183 - val_acc: 0.9031\n",
      "Epoch 116/150\n",
      "Epoch 00116: val_loss did not improve\n",
      " - 12s - loss: 0.1727 - acc: 0.9223 - val_loss: 0.2123 - val_acc: 0.9062\n",
      "Epoch 117/150\n",
      "Epoch 00117: val_loss did not improve\n",
      " - 12s - loss: 0.1745 - acc: 0.9237 - val_loss: 0.2320 - val_acc: 0.8969\n",
      "Epoch 118/150\n",
      "Epoch 00118: val_loss did not improve\n",
      " - 12s - loss: 0.1543 - acc: 0.9314 - val_loss: 0.2262 - val_acc: 0.8812\n",
      "Epoch 119/150\n",
      "Epoch 00119: val_loss did not improve\n",
      " - 12s - loss: 0.1571 - acc: 0.9330 - val_loss: 0.2222 - val_acc: 0.8938\n",
      "Epoch 120/150\n",
      "Epoch 00120: val_loss did not improve\n",
      " - 12s - loss: 0.1832 - acc: 0.9123 - val_loss: 0.2103 - val_acc: 0.9031\n",
      "Epoch 121/150\n",
      "Epoch 00121: val_loss did not improve\n",
      " - 12s - loss: 0.1636 - acc: 0.9306 - val_loss: 0.2239 - val_acc: 0.9125\n",
      "Epoch 122/150\n",
      "Epoch 00122: val_loss did not improve\n",
      " - 12s - loss: 0.1763 - acc: 0.9359 - val_loss: 0.2370 - val_acc: 0.8750\n",
      "Epoch 123/150\n",
      "Epoch 00123: val_loss did not improve\n",
      " - 12s - loss: 0.1592 - acc: 0.9298 - val_loss: 0.2102 - val_acc: 0.8938\n",
      "Epoch 124/150\n",
      "Epoch 00124: val_loss did not improve\n",
      " - 12s - loss: 0.1459 - acc: 0.9398 - val_loss: 0.2265 - val_acc: 0.8938\n",
      "Epoch 125/150\n",
      "Epoch 00125: val_loss did not improve\n",
      " - 12s - loss: 0.2013 - acc: 0.9011 - val_loss: 0.2831 - val_acc: 0.8562\n",
      "Epoch 126/150\n",
      "Epoch 00126: val_loss did not improve\n",
      " - 12s - loss: 0.1964 - acc: 0.9222 - val_loss: 0.2366 - val_acc: 0.9031\n",
      "Epoch 127/150\n",
      "Epoch 00127: val_loss did not improve\n",
      " - 12s - loss: 0.1673 - acc: 0.9231 - val_loss: 0.2291 - val_acc: 0.9031\n",
      "Epoch 128/150\n",
      "Epoch 00128: val_loss did not improve\n",
      " - 12s - loss: 0.1752 - acc: 0.9207 - val_loss: 0.2297 - val_acc: 0.8875\n",
      "Epoch 129/150\n",
      "Epoch 00129: val_loss did not improve\n",
      " - 12s - loss: 0.1710 - acc: 0.9306 - val_loss: 0.2201 - val_acc: 0.9031\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 130/150\n",
      "Epoch 00130: val_loss did not improve\n",
      " - 12s - loss: 0.1478 - acc: 0.9337 - val_loss: 0.2433 - val_acc: 0.9125\n",
      "Epoch 131/150\n",
      "Epoch 00131: val_loss did not improve\n",
      " - 12s - loss: 0.1601 - acc: 0.9314 - val_loss: 0.2198 - val_acc: 0.8969\n",
      "Epoch 132/150\n",
      "Epoch 00132: val_loss did not improve\n",
      " - 12s - loss: 0.1546 - acc: 0.9337 - val_loss: 0.2439 - val_acc: 0.9000\n",
      "Epoch 133/150\n",
      "Epoch 00133: val_loss did not improve\n",
      " - 12s - loss: 0.1597 - acc: 0.9298 - val_loss: 0.2368 - val_acc: 0.8844\n",
      "Epoch 134/150\n",
      "Epoch 00134: val_loss did not improve\n",
      " - 12s - loss: 0.1513 - acc: 0.9321 - val_loss: 0.2262 - val_acc: 0.9000\n",
      "Epoch 135/150\n",
      "Epoch 00135: val_loss did not improve\n",
      " - 12s - loss: 0.1358 - acc: 0.9436 - val_loss: 0.2208 - val_acc: 0.8969\n",
      "Epoch 136/150\n",
      "Epoch 00136: val_loss did not improve\n",
      " - 12s - loss: 0.1638 - acc: 0.9390 - val_loss: 0.2189 - val_acc: 0.8969\n",
      "Epoch 137/150\n",
      "Epoch 00137: val_loss did not improve\n",
      " - 12s - loss: 0.1459 - acc: 0.9451 - val_loss: 0.2258 - val_acc: 0.8969\n",
      "Epoch 138/150\n",
      "Epoch 00138: val_loss did not improve\n",
      " - 12s - loss: 0.1474 - acc: 0.9382 - val_loss: 0.2208 - val_acc: 0.9000\n",
      "Epoch 139/150\n",
      "Epoch 00139: val_loss did not improve\n",
      " - 12s - loss: 0.1525 - acc: 0.9307 - val_loss: 0.2339 - val_acc: 0.8844\n",
      "Epoch 140/150\n",
      "Epoch 00140: val_loss did not improve\n",
      " - 12s - loss: 0.1501 - acc: 0.9352 - val_loss: 0.2598 - val_acc: 0.8875\n",
      "Epoch 141/150\n",
      "Epoch 00141: val_loss did not improve\n",
      " - 12s - loss: 0.1541 - acc: 0.9315 - val_loss: 0.3247 - val_acc: 0.8906\n",
      "Epoch 142/150\n",
      "Epoch 00142: val_loss did not improve\n",
      " - 12s - loss: 0.1531 - acc: 0.9298 - val_loss: 0.2206 - val_acc: 0.8969\n",
      "Epoch 143/150\n",
      "Epoch 00143: val_loss did not improve\n",
      " - 12s - loss: 0.1524 - acc: 0.9359 - val_loss: 0.2194 - val_acc: 0.9125\n",
      "Epoch 144/150\n",
      "Epoch 00144: val_loss did not improve\n",
      " - 12s - loss: 0.1562 - acc: 0.9321 - val_loss: 0.2190 - val_acc: 0.9000\n",
      "Satge 2,done!\n",
      "1284/1284 [==============================] - 3s 3ms/step\n",
      "Train loss: 0.14215016321\n",
      "Train accuracy: 0.941588785047\n",
      "320/320 [==============================] - 1s 3ms/step\n",
      "Valid loss: 0.197189450264\n",
      "Valid accuracy: 0.915625\n",
      "\n",
      "===================FOLD= 5\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_5 (InputLayer)         (None, 75, 75, 3)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_17 (Conv2D)           (None, 71, 71, 128)       9728      \n",
      "_________________________________________________________________\n",
      "activation_17 (Activation)   (None, 71, 71, 128)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_17 (MaxPooling (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "dropout_25 (Dropout)         (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_18 (Conv2D)           (None, 33, 33, 256)       295168    \n",
      "_________________________________________________________________\n",
      "activation_18 (Activation)   (None, 33, 33, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_18 (MaxPooling (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "dropout_26 (Dropout)         (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_19 (Conv2D)           (None, 14, 14, 256)       590080    \n",
      "_________________________________________________________________\n",
      "activation_19 (Activation)   (None, 14, 14, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_19 (MaxPooling (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "dropout_27 (Dropout)         (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_20 (Conv2D)           (None, 5, 5, 128)         295040    \n",
      "_________________________________________________________________\n",
      "activation_20 (Activation)   (None, 5, 5, 128)         0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_20 (MaxPooling (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "dropout_28 (Dropout)         (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten_5 (Flatten)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_13 (Dense)             (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dropout_29 (Dropout)         (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_14 (Dense)             (None, 256)               131328    \n",
      "_________________________________________________________________\n",
      "dropout_30 (Dropout)         (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_15 (Dense)             (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 1,584,257\n",
      "Trainable params: 1,584,257\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/150\n",
      "Epoch 00001: val_loss improved from inf to 0.68611, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 14s - loss: 1.1453 - acc: 0.5153 - val_loss: 0.6861 - val_acc: 0.6062\n",
      "Epoch 2/150\n",
      "Epoch 00002: val_loss improved from 0.68611 to 0.65924, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.7341 - acc: 0.5253 - val_loss: 0.6592 - val_acc: 0.5594\n",
      "Epoch 3/150\n",
      "Epoch 00003: val_loss improved from 0.65924 to 0.59486, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.6435 - acc: 0.5877 - val_loss: 0.5949 - val_acc: 0.6625\n",
      "Epoch 4/150\n",
      "Epoch 00004: val_loss improved from 0.59486 to 0.57116, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.6009 - acc: 0.6060 - val_loss: 0.5712 - val_acc: 0.6625\n",
      "Epoch 5/150\n",
      "Epoch 00005: val_loss improved from 0.57116 to 0.56077, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.5671 - acc: 0.6492 - val_loss: 0.5608 - val_acc: 0.6781\n",
      "Epoch 6/150\n",
      "Epoch 00006: val_loss improved from 0.56077 to 0.55638, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.5630 - acc: 0.6479 - val_loss: 0.5564 - val_acc: 0.6875\n",
      "Epoch 7/150\n",
      "Epoch 00007: val_loss improved from 0.55638 to 0.54037, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.5357 - acc: 0.6988 - val_loss: 0.5404 - val_acc: 0.7000\n",
      "Epoch 8/150\n",
      "Epoch 00008: val_loss improved from 0.54037 to 0.50872, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.5155 - acc: 0.7135 - val_loss: 0.5087 - val_acc: 0.7344\n",
      "Epoch 9/150\n",
      "Epoch 00009: val_loss did not improve\n",
      " - 12s - loss: 0.5201 - acc: 0.7125 - val_loss: 0.5249 - val_acc: 0.7219\n",
      "Epoch 10/150\n",
      "Epoch 00010: val_loss improved from 0.50872 to 0.48135, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.4886 - acc: 0.7422 - val_loss: 0.4813 - val_acc: 0.7688\n",
      "Epoch 11/150\n",
      "Epoch 00011: val_loss improved from 0.48135 to 0.47707, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.4538 - acc: 0.7797 - val_loss: 0.4771 - val_acc: 0.7500\n",
      "Epoch 12/150\n",
      "Epoch 00012: val_loss improved from 0.47707 to 0.40929, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 13s - loss: 0.4609 - acc: 0.7754 - val_loss: 0.4093 - val_acc: 0.8125\n",
      "Epoch 13/150\n",
      "Epoch 00013: val_loss improved from 0.40929 to 0.40012, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.4178 - acc: 0.8087 - val_loss: 0.4001 - val_acc: 0.8250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 14/150\n",
      "Epoch 00014: val_loss improved from 0.40012 to 0.39917, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.4095 - acc: 0.8078 - val_loss: 0.3992 - val_acc: 0.8156\n",
      "Epoch 15/150\n",
      "Epoch 00015: val_loss improved from 0.39917 to 0.36036, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.3884 - acc: 0.8240 - val_loss: 0.3604 - val_acc: 0.8313\n",
      "Epoch 16/150\n",
      "Epoch 00016: val_loss improved from 0.36036 to 0.35390, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.3835 - acc: 0.8330 - val_loss: 0.3539 - val_acc: 0.8438\n",
      "Epoch 17/150\n",
      "Epoch 00017: val_loss did not improve\n",
      " - 12s - loss: 0.3674 - acc: 0.8376 - val_loss: 0.4173 - val_acc: 0.7969\n",
      "Epoch 18/150\n",
      "Epoch 00018: val_loss did not improve\n",
      " - 12s - loss: 0.3722 - acc: 0.8246 - val_loss: 0.3738 - val_acc: 0.8250\n",
      "Epoch 19/150\n",
      "Epoch 00019: val_loss improved from 0.35390 to 0.34481, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.3550 - acc: 0.8376 - val_loss: 0.3448 - val_acc: 0.8250\n",
      "Epoch 20/150\n",
      "Epoch 00020: val_loss did not improve\n",
      " - 12s - loss: 0.3480 - acc: 0.8429 - val_loss: 0.4111 - val_acc: 0.7844\n",
      "Epoch 21/150\n",
      "Epoch 00021: val_loss improved from 0.34481 to 0.32913, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.3602 - acc: 0.8354 - val_loss: 0.3291 - val_acc: 0.8562\n",
      "Epoch 22/150\n",
      "Epoch 00022: val_loss improved from 0.32913 to 0.31152, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.3318 - acc: 0.8583 - val_loss: 0.3115 - val_acc: 0.8594\n",
      "Epoch 23/150\n",
      "Epoch 00023: val_loss did not improve\n",
      " - 12s - loss: 0.3199 - acc: 0.8612 - val_loss: 0.3126 - val_acc: 0.8500\n",
      "Epoch 24/150\n",
      "Epoch 00024: val_loss did not improve\n",
      " - 12s - loss: 0.3424 - acc: 0.8414 - val_loss: 0.3584 - val_acc: 0.8219\n",
      "Epoch 25/150\n",
      "Epoch 00025: val_loss improved from 0.31152 to 0.29434, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.3266 - acc: 0.8543 - val_loss: 0.2943 - val_acc: 0.8625\n",
      "Epoch 26/150\n",
      "Epoch 00026: val_loss did not improve\n",
      " - 12s - loss: 0.3226 - acc: 0.8551 - val_loss: 0.3427 - val_acc: 0.8375\n",
      "Epoch 27/150\n",
      "Epoch 00027: val_loss did not improve\n",
      " - 12s - loss: 0.3066 - acc: 0.8604 - val_loss: 0.3004 - val_acc: 0.8313\n",
      "Epoch 28/150\n",
      "Epoch 00028: val_loss improved from 0.29434 to 0.29329, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.3053 - acc: 0.8604 - val_loss: 0.2933 - val_acc: 0.8688\n",
      "Epoch 29/150\n",
      "Epoch 00029: val_loss did not improve\n",
      " - 12s - loss: 0.3101 - acc: 0.8566 - val_loss: 0.3146 - val_acc: 0.8625\n",
      "Epoch 30/150\n",
      "Epoch 00030: val_loss did not improve\n",
      " - 12s - loss: 0.2945 - acc: 0.8606 - val_loss: 0.3147 - val_acc: 0.8375\n",
      "Epoch 31/150\n",
      "Epoch 00031: val_loss did not improve\n",
      " - 12s - loss: 0.2990 - acc: 0.8638 - val_loss: 0.2968 - val_acc: 0.8438\n",
      "Epoch 32/150\n",
      "Epoch 00032: val_loss did not improve\n",
      " - 12s - loss: 0.3024 - acc: 0.8537 - val_loss: 0.3000 - val_acc: 0.8719\n",
      "Epoch 33/150\n",
      "Epoch 00033: val_loss did not improve\n",
      " - 12s - loss: 0.3128 - acc: 0.8615 - val_loss: 0.3856 - val_acc: 0.7969\n",
      "Epoch 34/150\n",
      "Epoch 00034: val_loss improved from 0.29329 to 0.28061, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.3015 - acc: 0.8597 - val_loss: 0.2806 - val_acc: 0.8656\n",
      "Epoch 35/150\n",
      "Epoch 00035: val_loss improved from 0.28061 to 0.26381, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.3173 - acc: 0.8598 - val_loss: 0.2638 - val_acc: 0.8844\n",
      "Epoch 36/150\n",
      "Epoch 00036: val_loss did not improve\n",
      " - 12s - loss: 0.2798 - acc: 0.8787 - val_loss: 0.2693 - val_acc: 0.8812\n",
      "Epoch 37/150\n",
      "Epoch 00037: val_loss improved from 0.26381 to 0.25534, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.2678 - acc: 0.8902 - val_loss: 0.2553 - val_acc: 0.8875\n",
      "Epoch 38/150\n",
      "Epoch 00038: val_loss did not improve\n",
      " - 12s - loss: 0.3065 - acc: 0.8543 - val_loss: 0.2632 - val_acc: 0.8812\n",
      "Epoch 39/150\n",
      "Epoch 00039: val_loss improved from 0.25534 to 0.24772, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.2580 - acc: 0.8909 - val_loss: 0.2477 - val_acc: 0.8875\n",
      "Epoch 40/150\n",
      "Epoch 00040: val_loss did not improve\n",
      " - 12s - loss: 0.2838 - acc: 0.8781 - val_loss: 0.2636 - val_acc: 0.8781\n",
      "Epoch 41/150\n",
      "Epoch 00041: val_loss did not improve\n",
      " - 12s - loss: 0.2691 - acc: 0.8751 - val_loss: 0.2624 - val_acc: 0.8844\n",
      "Epoch 42/150\n",
      "Epoch 00042: val_loss did not improve\n",
      " - 12s - loss: 0.2595 - acc: 0.8864 - val_loss: 0.2502 - val_acc: 0.8906\n",
      "Epoch 43/150\n",
      "Epoch 00043: val_loss did not improve\n",
      " - 12s - loss: 0.2686 - acc: 0.8751 - val_loss: 0.2702 - val_acc: 0.8688\n",
      "Epoch 44/150\n",
      "Epoch 00044: val_loss did not improve\n",
      " - 12s - loss: 0.2734 - acc: 0.8810 - val_loss: 0.2570 - val_acc: 0.8812\n",
      "Epoch 45/150\n",
      "Epoch 00045: val_loss did not improve\n",
      " - 12s - loss: 0.2641 - acc: 0.8865 - val_loss: 0.3498 - val_acc: 0.8219\n",
      "Epoch 46/150\n",
      "Epoch 00046: val_loss did not improve\n",
      " - 12s - loss: 0.2727 - acc: 0.8810 - val_loss: 0.2611 - val_acc: 0.8750\n",
      "Epoch 47/150\n",
      "Epoch 00047: val_loss did not improve\n",
      " - 12s - loss: 0.2610 - acc: 0.8757 - val_loss: 0.2706 - val_acc: 0.8719\n",
      "Epoch 48/150\n",
      "Epoch 00048: val_loss did not improve\n",
      " - 12s - loss: 0.2542 - acc: 0.8857 - val_loss: 0.2533 - val_acc: 0.9062\n",
      "Epoch 49/150\n",
      "Epoch 00049: val_loss did not improve\n",
      " - 12s - loss: 0.2410 - acc: 0.8909 - val_loss: 0.2520 - val_acc: 0.8844\n",
      "Epoch 50/150\n",
      "Epoch 00050: val_loss improved from 0.24772 to 0.24065, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.2454 - acc: 0.8864 - val_loss: 0.2406 - val_acc: 0.9000\n",
      "Epoch 51/150\n",
      "Epoch 00051: val_loss did not improve\n",
      " - 12s - loss: 0.2482 - acc: 0.8835 - val_loss: 0.2458 - val_acc: 0.8969\n",
      "Epoch 52/150\n",
      "Epoch 00052: val_loss improved from 0.24065 to 0.23354, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.2665 - acc: 0.8751 - val_loss: 0.2335 - val_acc: 0.9031\n",
      "Epoch 53/150\n",
      "Epoch 00053: val_loss improved from 0.23354 to 0.23279, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.2449 - acc: 0.8873 - val_loss: 0.2328 - val_acc: 0.9094\n",
      "Epoch 54/150\n",
      "Epoch 00054: val_loss did not improve\n",
      " - 12s - loss: 0.2536 - acc: 0.8873 - val_loss: 0.2461 - val_acc: 0.8969\n",
      "Epoch 55/150\n",
      "Epoch 00055: val_loss improved from 0.23279 to 0.22643, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.2373 - acc: 0.8850 - val_loss: 0.2264 - val_acc: 0.9062\n",
      "Epoch 56/150\n",
      "Epoch 00056: val_loss did not improve\n",
      " - 12s - loss: 0.2228 - acc: 0.9100 - val_loss: 0.2292 - val_acc: 0.9062\n",
      "Epoch 57/150\n",
      "Epoch 00057: val_loss did not improve\n",
      " - 12s - loss: 0.2467 - acc: 0.8888 - val_loss: 0.2378 - val_acc: 0.8906\n",
      "Epoch 58/150\n",
      "Epoch 00058: val_loss did not improve\n",
      " - 12s - loss: 0.2423 - acc: 0.8902 - val_loss: 0.2715 - val_acc: 0.8719\n",
      "Epoch 59/150\n",
      "Epoch 00059: val_loss did not improve\n",
      " - 12s - loss: 0.2445 - acc: 0.8940 - val_loss: 0.2595 - val_acc: 0.8875\n",
      "Epoch 60/150\n",
      "Epoch 00060: val_loss did not improve\n",
      " - 12s - loss: 0.2437 - acc: 0.8804 - val_loss: 0.2414 - val_acc: 0.9062\n",
      "Epoch 61/150\n",
      "Epoch 00061: val_loss did not improve\n",
      " - 12s - loss: 0.2346 - acc: 0.9047 - val_loss: 0.2438 - val_acc: 0.8938\n",
      "Epoch 62/150\n",
      "Epoch 00062: val_loss did not improve\n",
      " - 12s - loss: 0.2461 - acc: 0.8888 - val_loss: 0.2423 - val_acc: 0.8906\n",
      "Epoch 63/150\n",
      "Epoch 00063: val_loss improved from 0.22643 to 0.22608, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.2309 - acc: 0.8978 - val_loss: 0.2261 - val_acc: 0.9094\n",
      "Epoch 64/150\n",
      "Epoch 00064: val_loss did not improve\n",
      " - 12s - loss: 0.2084 - acc: 0.9062 - val_loss: 0.2541 - val_acc: 0.8969\n",
      "Epoch 65/150\n",
      "Epoch 00065: val_loss did not improve\n",
      " - 12s - loss: 0.2244 - acc: 0.8978 - val_loss: 0.2350 - val_acc: 0.9094\n",
      "Epoch 66/150\n",
      "Epoch 00066: val_loss did not improve\n",
      " - 12s - loss: 0.2302 - acc: 0.8986 - val_loss: 0.2562 - val_acc: 0.9094\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 67/150\n",
      "Epoch 00067: val_loss did not improve\n",
      " - 12s - loss: 0.2440 - acc: 0.8903 - val_loss: 0.2379 - val_acc: 0.9094\n",
      "Epoch 68/150\n",
      "Epoch 00068: val_loss did not improve\n",
      " - 12s - loss: 0.2168 - acc: 0.9062 - val_loss: 0.2384 - val_acc: 0.9125\n",
      "Epoch 69/150\n",
      "Epoch 00069: val_loss did not improve\n",
      " - 12s - loss: 0.2357 - acc: 0.8932 - val_loss: 0.2538 - val_acc: 0.8969\n",
      "Epoch 70/150\n",
      "Epoch 00070: val_loss improved from 0.22608 to 0.22396, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.2200 - acc: 0.9077 - val_loss: 0.2240 - val_acc: 0.9062\n",
      "Epoch 71/150\n",
      "Epoch 00071: val_loss did not improve\n",
      " - 12s - loss: 0.2218 - acc: 0.8949 - val_loss: 0.2995 - val_acc: 0.8750\n",
      "Epoch 72/150\n",
      "Epoch 00072: val_loss improved from 0.22396 to 0.21565, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.2153 - acc: 0.9039 - val_loss: 0.2156 - val_acc: 0.9219\n",
      "Epoch 73/150\n",
      "Epoch 00073: val_loss did not improve\n",
      " - 12s - loss: 0.2186 - acc: 0.9032 - val_loss: 0.2161 - val_acc: 0.9156\n",
      "Epoch 74/150\n",
      "Epoch 00074: val_loss did not improve\n",
      " - 12s - loss: 0.2148 - acc: 0.9093 - val_loss: 0.2207 - val_acc: 0.9062\n",
      "Epoch 75/150\n",
      "Epoch 00075: val_loss did not improve\n",
      " - 12s - loss: 0.2199 - acc: 0.9070 - val_loss: 0.2334 - val_acc: 0.9156\n",
      "Epoch 76/150\n",
      "Epoch 00076: val_loss did not improve\n",
      " - 12s - loss: 0.2146 - acc: 0.9025 - val_loss: 0.2484 - val_acc: 0.9000\n",
      "Epoch 77/150\n",
      "Epoch 00077: val_loss did not improve\n",
      " - 12s - loss: 0.2443 - acc: 0.8911 - val_loss: 0.2911 - val_acc: 0.8344\n",
      "Epoch 78/150\n",
      "Epoch 00078: val_loss did not improve\n",
      " - 12s - loss: 0.2081 - acc: 0.9115 - val_loss: 0.2423 - val_acc: 0.9094\n",
      "Epoch 79/150\n",
      "Epoch 00079: val_loss did not improve\n",
      " - 12s - loss: 0.2329 - acc: 0.8925 - val_loss: 0.2445 - val_acc: 0.9000\n",
      "Epoch 80/150\n",
      "Epoch 00080: val_loss did not improve\n",
      " - 12s - loss: 0.2047 - acc: 0.9077 - val_loss: 0.2326 - val_acc: 0.9062\n",
      "Epoch 81/150\n",
      "Epoch 00081: val_loss did not improve\n",
      " - 12s - loss: 0.1997 - acc: 0.9085 - val_loss: 0.2311 - val_acc: 0.9000\n",
      "Epoch 82/150\n",
      "Epoch 00082: val_loss did not improve\n",
      " - 12s - loss: 0.2044 - acc: 0.9146 - val_loss: 0.2250 - val_acc: 0.9313\n",
      "Epoch 83/150\n",
      "Epoch 00083: val_loss did not improve\n",
      " - 12s - loss: 0.2088 - acc: 0.9033 - val_loss: 0.2452 - val_acc: 0.9125\n",
      "Epoch 84/150\n",
      "Epoch 00084: val_loss did not improve\n",
      " - 12s - loss: 0.2020 - acc: 0.9093 - val_loss: 0.2272 - val_acc: 0.9031\n",
      "Epoch 85/150\n",
      "Epoch 00085: val_loss did not improve\n",
      " - 12s - loss: 0.2040 - acc: 0.9169 - val_loss: 0.2262 - val_acc: 0.9219\n",
      "Epoch 86/150\n",
      "Epoch 00086: val_loss did not improve\n",
      " - 12s - loss: 0.2361 - acc: 0.8896 - val_loss: 0.2412 - val_acc: 0.8938\n",
      "Epoch 87/150\n",
      "Epoch 00087: val_loss did not improve\n",
      " - 12s - loss: 0.2106 - acc: 0.9024 - val_loss: 0.2294 - val_acc: 0.8969\n",
      "Epoch 88/150\n",
      "Epoch 00088: val_loss did not improve\n",
      " - 12s - loss: 0.2083 - acc: 0.9131 - val_loss: 0.2307 - val_acc: 0.9219\n",
      "Epoch 89/150\n",
      "Epoch 00089: val_loss did not improve\n",
      " - 12s - loss: 0.1885 - acc: 0.9192 - val_loss: 0.2199 - val_acc: 0.9156\n",
      "Epoch 90/150\n",
      "Epoch 00090: val_loss did not improve\n",
      " - 12s - loss: 0.1893 - acc: 0.9176 - val_loss: 0.2666 - val_acc: 0.9062\n",
      "Epoch 91/150\n",
      "Epoch 00091: val_loss did not improve\n",
      " - 12s - loss: 0.2268 - acc: 0.9047 - val_loss: 0.2174 - val_acc: 0.9125\n",
      "Epoch 92/150\n",
      "Epoch 00092: val_loss did not improve\n",
      " - 12s - loss: 0.1947 - acc: 0.9147 - val_loss: 0.2167 - val_acc: 0.9156\n",
      "Epoch 93/150\n",
      "Epoch 00093: val_loss did not improve\n",
      " - 12s - loss: 0.1967 - acc: 0.9131 - val_loss: 0.2583 - val_acc: 0.9125\n",
      "Epoch 94/150\n",
      "Epoch 00094: val_loss did not improve\n",
      " - 12s - loss: 0.2338 - acc: 0.9018 - val_loss: 0.2352 - val_acc: 0.9062\n",
      "Epoch 95/150\n",
      "Epoch 00095: val_loss did not improve\n",
      " - 12s - loss: 0.1984 - acc: 0.9109 - val_loss: 0.2327 - val_acc: 0.9187\n",
      "Epoch 96/150\n",
      "Epoch 00096: val_loss improved from 0.21565 to 0.21289, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.2043 - acc: 0.9048 - val_loss: 0.2129 - val_acc: 0.9219\n",
      "Epoch 97/150\n",
      "Epoch 00097: val_loss did not improve\n",
      " - 12s - loss: 0.1949 - acc: 0.9077 - val_loss: 0.2199 - val_acc: 0.9250\n",
      "Epoch 98/150\n",
      "Epoch 00098: val_loss did not improve\n",
      " - 12s - loss: 0.1959 - acc: 0.9161 - val_loss: 0.2422 - val_acc: 0.9094\n",
      "Epoch 99/150\n",
      "Epoch 00099: val_loss did not improve\n",
      " - 12s - loss: 0.1877 - acc: 0.9176 - val_loss: 0.2245 - val_acc: 0.9062\n",
      "Epoch 100/150\n",
      "Epoch 00100: val_loss did not improve\n",
      " - 12s - loss: 0.1768 - acc: 0.9192 - val_loss: 0.2221 - val_acc: 0.9219\n",
      "Epoch 101/150\n",
      "Epoch 00101: val_loss did not improve\n",
      " - 12s - loss: 0.1876 - acc: 0.9199 - val_loss: 0.2197 - val_acc: 0.9031\n",
      "Epoch 102/150\n",
      "Epoch 00102: val_loss did not improve\n",
      " - 12s - loss: 0.1922 - acc: 0.9123 - val_loss: 0.2158 - val_acc: 0.9250\n",
      "Epoch 103/150\n",
      "Epoch 00103: val_loss did not improve\n",
      " - 12s - loss: 0.1786 - acc: 0.9161 - val_loss: 0.2141 - val_acc: 0.9125\n",
      "Epoch 104/150\n",
      "Epoch 00104: val_loss improved from 0.21289 to 0.20843, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.1928 - acc: 0.9093 - val_loss: 0.2084 - val_acc: 0.9250\n",
      "Epoch 105/150\n",
      "Epoch 00105: val_loss did not improve\n",
      " - 12s - loss: 0.2198 - acc: 0.9010 - val_loss: 0.2269 - val_acc: 0.9125\n",
      "Epoch 106/150\n",
      "Epoch 00106: val_loss did not improve\n",
      " - 12s - loss: 0.1785 - acc: 0.9230 - val_loss: 0.2587 - val_acc: 0.9062\n",
      "Epoch 107/150\n",
      "Epoch 00107: val_loss did not improve\n",
      " - 12s - loss: 0.1778 - acc: 0.9306 - val_loss: 0.2209 - val_acc: 0.9094\n",
      "Epoch 108/150\n",
      "Epoch 00108: val_loss did not improve\n",
      " - 12s - loss: 0.1957 - acc: 0.9146 - val_loss: 0.2234 - val_acc: 0.9187\n",
      "Epoch 109/150\n",
      "Epoch 00109: val_loss did not improve\n",
      " - 12s - loss: 0.1834 - acc: 0.9230 - val_loss: 0.2333 - val_acc: 0.9250\n",
      "Epoch 110/150\n",
      "Epoch 00110: val_loss did not improve\n",
      " - 12s - loss: 0.1765 - acc: 0.9306 - val_loss: 0.2166 - val_acc: 0.9156\n",
      "Epoch 111/150\n",
      "Epoch 00111: val_loss did not improve\n",
      " - 12s - loss: 0.1746 - acc: 0.9230 - val_loss: 0.2146 - val_acc: 0.9313\n",
      "Epoch 112/150\n",
      "Epoch 00112: val_loss did not improve\n",
      " - 12s - loss: 0.1806 - acc: 0.9184 - val_loss: 0.2208 - val_acc: 0.9250\n",
      "Epoch 113/150\n",
      "Epoch 00113: val_loss improved from 0.20843 to 0.20811, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.1931 - acc: 0.9169 - val_loss: 0.2081 - val_acc: 0.9375\n",
      "Epoch 114/150\n",
      "Epoch 00114: val_loss did not improve\n",
      " - 12s - loss: 0.2190 - acc: 0.8952 - val_loss: 0.2279 - val_acc: 0.9125\n",
      "Epoch 115/150\n",
      "Epoch 00115: val_loss did not improve\n",
      " - 12s - loss: 0.1811 - acc: 0.9199 - val_loss: 0.2307 - val_acc: 0.9219\n",
      "Epoch 116/150\n",
      "Epoch 00116: val_loss did not improve\n",
      " - 12s - loss: 0.1831 - acc: 0.9207 - val_loss: 0.2447 - val_acc: 0.9000\n",
      "Epoch 117/150\n",
      "Epoch 00117: val_loss did not improve\n",
      " - 12s - loss: 0.1673 - acc: 0.9337 - val_loss: 0.2309 - val_acc: 0.9187\n",
      "Epoch 118/150\n",
      "Epoch 00118: val_loss did not improve\n",
      " - 12s - loss: 0.1738 - acc: 0.9268 - val_loss: 0.2192 - val_acc: 0.9187\n",
      "Epoch 119/150\n",
      "Epoch 00119: val_loss did not improve\n",
      " - 12s - loss: 0.1780 - acc: 0.9192 - val_loss: 0.2111 - val_acc: 0.9219\n",
      "Epoch 120/150\n",
      "Epoch 00120: val_loss did not improve\n",
      " - 12s - loss: 0.1770 - acc: 0.9178 - val_loss: 0.2100 - val_acc: 0.9281\n",
      "Epoch 121/150\n",
      "Epoch 00121: val_loss improved from 0.20811 to 0.20553, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.1735 - acc: 0.9176 - val_loss: 0.2055 - val_acc: 0.9281\n",
      "Epoch 122/150\n",
      "Epoch 00122: val_loss did not improve\n",
      " - 12s - loss: 0.1633 - acc: 0.9298 - val_loss: 0.2295 - val_acc: 0.9187\n",
      "Epoch 123/150\n",
      "Epoch 00123: val_loss improved from 0.20553 to 0.20425, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.1752 - acc: 0.9352 - val_loss: 0.2042 - val_acc: 0.9313\n",
      "Epoch 124/150\n",
      "Epoch 00124: val_loss did not improve\n",
      " - 12s - loss: 0.1711 - acc: 0.9260 - val_loss: 0.2047 - val_acc: 0.9375\n",
      "Epoch 125/150\n",
      "Epoch 00125: val_loss did not improve\n",
      " - 12s - loss: 0.1816 - acc: 0.9193 - val_loss: 0.2350 - val_acc: 0.9094\n",
      "Epoch 126/150\n",
      "Epoch 00126: val_loss did not improve\n",
      " - 12s - loss: 0.1944 - acc: 0.9123 - val_loss: 0.2269 - val_acc: 0.9031\n",
      "Epoch 127/150\n",
      "Epoch 00127: val_loss did not improve\n",
      " - 12s - loss: 0.1596 - acc: 0.9321 - val_loss: 0.2073 - val_acc: 0.9219\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 128/150\n",
      "Epoch 00128: val_loss improved from 0.20425 to 0.20352, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.1497 - acc: 0.9375 - val_loss: 0.2035 - val_acc: 0.9219\n",
      "Epoch 129/150\n",
      "Epoch 00129: val_loss did not improve\n",
      " - 12s - loss: 0.1665 - acc: 0.9260 - val_loss: 0.2157 - val_acc: 0.9125\n",
      "Epoch 130/150\n",
      "Epoch 00130: val_loss did not improve\n",
      " - 12s - loss: 0.1527 - acc: 0.9337 - val_loss: 0.2086 - val_acc: 0.9062\n",
      "Epoch 131/150\n",
      "Epoch 00131: val_loss did not improve\n",
      " - 12s - loss: 0.1464 - acc: 0.9405 - val_loss: 0.2139 - val_acc: 0.9125\n",
      "Epoch 132/150\n",
      "Epoch 00132: val_loss did not improve\n",
      " - 12s - loss: 0.1595 - acc: 0.9306 - val_loss: 0.2316 - val_acc: 0.9125\n",
      "Epoch 133/150\n",
      "Epoch 00133: val_loss did not improve\n",
      " - 12s - loss: 0.1753 - acc: 0.9154 - val_loss: 0.2339 - val_acc: 0.9125\n",
      "Epoch 134/150\n",
      "Epoch 00134: val_loss did not improve\n",
      " - 12s - loss: 0.1579 - acc: 0.9254 - val_loss: 0.2076 - val_acc: 0.9156\n",
      "Epoch 135/150\n",
      "Epoch 00135: val_loss did not improve\n",
      " - 12s - loss: 0.1673 - acc: 0.9231 - val_loss: 0.2159 - val_acc: 0.9125\n",
      "Epoch 136/150\n",
      "Epoch 00136: val_loss improved from 0.20352 to 0.19682, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.1579 - acc: 0.9230 - val_loss: 0.1968 - val_acc: 0.9344\n",
      "Epoch 137/150\n",
      "Epoch 00137: val_loss did not improve\n",
      " - 12s - loss: 0.1604 - acc: 0.9344 - val_loss: 0.2914 - val_acc: 0.8906\n",
      "Epoch 138/150\n",
      "Epoch 00138: val_loss did not improve\n",
      " - 12s - loss: 0.1424 - acc: 0.9466 - val_loss: 0.2224 - val_acc: 0.9281\n",
      "Epoch 139/150\n",
      "Epoch 00139: val_loss did not improve\n",
      " - 12s - loss: 0.1640 - acc: 0.9321 - val_loss: 0.2043 - val_acc: 0.9281\n",
      "Epoch 140/150\n",
      "Epoch 00140: val_loss did not improve\n",
      " - 12s - loss: 0.1654 - acc: 0.9223 - val_loss: 0.2033 - val_acc: 0.9094\n",
      "Epoch 141/150\n",
      "Epoch 00141: val_loss improved from 0.19682 to 0.19422, saving model to model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv5.hdf5\n",
      " - 12s - loss: 0.1639 - acc: 0.9300 - val_loss: 0.1942 - val_acc: 0.9344\n",
      "Epoch 142/150\n",
      "Epoch 00142: val_loss did not improve\n",
      " - 12s - loss: 0.1546 - acc: 0.9329 - val_loss: 0.2594 - val_acc: 0.9094\n",
      "Epoch 143/150\n",
      "Epoch 00143: val_loss did not improve\n",
      " - 12s - loss: 0.1567 - acc: 0.9260 - val_loss: 0.2106 - val_acc: 0.9281\n",
      "Epoch 144/150\n",
      "Epoch 00144: val_loss did not improve\n",
      " - 12s - loss: 0.1456 - acc: 0.9359 - val_loss: 0.1956 - val_acc: 0.9313\n",
      "Epoch 145/150\n",
      "Epoch 00145: val_loss did not improve\n",
      " - 12s - loss: 0.1832 - acc: 0.9384 - val_loss: 0.2011 - val_acc: 0.9313\n",
      "Epoch 146/150\n",
      "Epoch 00146: val_loss did not improve\n",
      " - 12s - loss: 0.1627 - acc: 0.9284 - val_loss: 0.2517 - val_acc: 0.9031\n",
      "Epoch 147/150\n",
      "Epoch 00147: val_loss did not improve\n",
      " - 12s - loss: 0.1649 - acc: 0.9337 - val_loss: 0.1996 - val_acc: 0.9344\n",
      "Epoch 148/150\n",
      "Epoch 00148: val_loss did not improve\n",
      " - 12s - loss: 0.1425 - acc: 0.9443 - val_loss: 0.2110 - val_acc: 0.9156\n",
      "Epoch 149/150\n",
      "Epoch 00149: val_loss did not improve\n",
      " - 12s - loss: 0.1466 - acc: 0.9367 - val_loss: 0.2024 - val_acc: 0.9219\n",
      "Epoch 150/150\n",
      "Epoch 00150: val_loss did not improve\n",
      " - 12s - loss: 0.2033 - acc: 0.9216 - val_loss: 0.2222 - val_acc: 0.8844\n",
      "Satge 2,done!\n",
      "1284/1284 [==============================] - 4s 3ms/step\n",
      "Train loss: 0.138621065197\n",
      "Train accuracy: 0.939252336449\n",
      "320/320 [==============================] - 1s 3ms/step\n",
      "Valid loss: 0.194219503179\n",
      "Valid accuracy: 0.934375\n",
      "\n",
      " Train Log Loss Validation=  0.158464488224\n",
      " Test Log Loss Validation=  0.215773077865\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0       2.634234e-05\n",
       "1       7.572710e-12\n",
       "2       8.207796e-01\n",
       "3       1.029958e-04\n",
       "4       1.034867e-02\n",
       "5       1.000000e+00\n",
       "6       9.873828e-01\n",
       "7       8.719710e-01\n",
       "8       1.457550e-11\n",
       "9       1.286021e-01\n",
       "10      3.451467e-01\n",
       "11      2.154507e-04\n",
       "12      9.107782e-01\n",
       "13      8.817458e-01\n",
       "14      4.322701e-03\n",
       "15      5.364039e-02\n",
       "16      5.240228e-01\n",
       "17      6.339528e-03\n",
       "18      4.478098e-01\n",
       "19      5.596439e-01\n",
       "20      2.318110e-02\n",
       "21      3.361613e-01\n",
       "22      4.370111e-08\n",
       "23      9.999779e-01\n",
       "24      7.067740e-09\n",
       "25      9.999441e-01\n",
       "26      9.404380e-02\n",
       "27      1.573823e-03\n",
       "28      5.041999e-01\n",
       "29      7.440081e-06\n",
       "            ...     \n",
       "1574    2.094985e-01\n",
       "1575    2.481993e-10\n",
       "1576    1.271278e-07\n",
       "1577    2.172237e-02\n",
       "1578    6.413934e-02\n",
       "1579    2.686754e-01\n",
       "1580    5.664628e-02\n",
       "1581    7.319692e-02\n",
       "1582    5.183475e-01\n",
       "1583    5.260556e-01\n",
       "1584    1.192241e-04\n",
       "1585    1.654384e-02\n",
       "1586    5.180713e-01\n",
       "1587    7.185864e-02\n",
       "1588    5.936407e-01\n",
       "1589    6.612022e-03\n",
       "1590    1.654102e-04\n",
       "1591    7.601394e-01\n",
       "1592    1.948547e-02\n",
       "1593    3.624265e-01\n",
       "1594    6.944709e-05\n",
       "1595    2.989503e-05\n",
       "1596    1.100085e-02\n",
       "1597    3.762044e-05\n",
       "1598    3.627428e-02\n",
       "1599    2.725236e-03\n",
       "1600    4.127137e-04\n",
       "1601    1.268125e-01\n",
       "1602    2.564098e-01\n",
       "1603    1.253185e-01\n",
       "Name: is_iceberg, Length: 1604, dtype: float64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Model7_CV(X_train,K = 5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Submission"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Evaluation_CV(X_train,K=3):\n",
    "    folds = list(StratifiedKFold(n_splits=K, shuffle=True,random_state = 3).split(X_train, target_train))\n",
    "    y_test_pred_log=0\n",
    "    y_train_pred_log=0\n",
    "    y_valid_pred_log = 0.0*target_train\n",
    "    for j, (train_idx, valid_idx) in enumerate(folds):\n",
    "        print('\\n===================FOLD=',j+1)\n",
    "        X_train_cv = X_train[train_idx]\n",
    "        y_train_cv = target_train[train_idx]\n",
    "        X_valid = X_train[valid_idx]\n",
    "        y_valid= target_train[valid_idx]\n",
    "        \n",
    "        #Angle\n",
    "#         X_angle_cv=X_angle[train_idx]\n",
    "#         X_angle_hold=X_angle[test_idx]\n",
    "\n",
    "        #define file path and get callbacks\n",
    "        file_path = 'model save/Model 6-Advanced CNN with DA and self TL(cv3)/Model-6 cv'+str(j+1)+'.hdf5'\n",
    "        Cnnmodel=cnnmodel(input_shape = (75,75,3),lr = 0.0001)\n",
    "        Cnnmodel.load_weights(file_path)\n",
    "\n",
    "\n",
    "#         callbacks = get_callbacks(filepath=file_path, patience=5)\n",
    "#         gen_flow = gen_flow_for_two_inputs(X_train_cv, X_angle_cv, y_train_cv)\n",
    "#         galaxyModel= getVggAngleModel()\n",
    "#         galaxyModel.fit_generator(\n",
    "#                 gen_flow,\n",
    "#                 steps_per_epoch=24,\n",
    "#                 epochs=100,\n",
    "#                 shuffle=True,\n",
    "#                 verbose=1,\n",
    "#                 validation_data=([X_holdout,X_angle_hold], Y_holdout),\n",
    "#                 callbacks=callbacks)\n",
    "\n",
    "        #Getting the Best Model\n",
    "        #Getting Training Score\n",
    "        print('cv'+str(j+1))\n",
    "        score = Cnnmodel.evaluate(X_train_cv, y_train_cv)\n",
    "        print('Train loss:', score[0])\n",
    "        print('Train accuracy:', score[1])\n",
    "        #Getting Test Score\n",
    "        score = Cnnmodel.evaluate(X_valid, y_valid)\n",
    "        print('Valid loss:', score[0])\n",
    "        print('Valid accuracy:', score[1])\n",
    "        print('\\n')\n",
    "\n",
    "        #Getting validation Score.\n",
    "        pred_valid=Cnnmodel.predict(X_valid)\n",
    "        y_valid_pred_log[valid_idx] = pred_valid.reshape(pred_valid.shape[0])\n",
    "\n",
    "        #Getting Test Scores\n",
    "        temp_test=Cnnmodel.predict(X_test)\n",
    "        y_test_pred_log+=temp_test.reshape(temp_test.shape[0])\n",
    "\n",
    "        #Getting Train Scores\n",
    "        temp_train=Cnnmodel.predict(X_train)\n",
    "        y_train_pred_log+=temp_train.reshape(temp_train.shape[0])\n",
    "\n",
    "    y_valid_pred_log=y_valid_pred_log/K\n",
    "    y_train_pred_log=y_train_pred_log/K\n",
    "\n",
    "    print('\\n Train Log Loss Validation= ',log_loss(target_train, y_train_pred_log))\n",
    "    print(' Test Log Loss Validation= ',log_loss(target_train, y_valid_pred_log))\n",
    "    return y_test_pred_log"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "preds = Evaluation_CV(X_train,K=5)\n",
    "sub = test[['id']]\n",
    "sub['is_iceberg'] = preds\n",
    "sub.to_csv('5cv.csv',index = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Submission(min max median stacking)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def test_scores(K=5):\n",
    "    y_test_pred_log=0\n",
    "    test_df = test_data[['id']]\n",
    "    for j in range(K):\n",
    "        print('\\n===================FOLD=',j+1)\n",
    "        file_path = 'model save/Model 8-Advanced CNN with DA(cv5)/Model-8 cv'+str(j+1)+'.hdf5'\n",
    "        Cnnmodel = model_compile(cnnmodel(),lr = 0.0001,decay = 0, freezing_layers = None,weights_path = file_path)\n",
    "\n",
    "        #Getting Test Scores\n",
    "        temp_test=Cnnmodel.predict(X_test)\n",
    "        test_df['cv'+str(j+1)]=temp_test.reshape(temp_test.shape[0])\n",
    "    return test_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "===================FOLD= 1\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_6 (InputLayer)         (None, 75, 75, 3)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_21 (Conv2D)           (None, 71, 71, 128)       9728      \n",
      "_________________________________________________________________\n",
      "activation_21 (Activation)   (None, 71, 71, 128)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_21 (MaxPooling (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "dropout_31 (Dropout)         (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_22 (Conv2D)           (None, 33, 33, 256)       295168    \n",
      "_________________________________________________________________\n",
      "activation_22 (Activation)   (None, 33, 33, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_22 (MaxPooling (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "dropout_32 (Dropout)         (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_23 (Conv2D)           (None, 14, 14, 256)       590080    \n",
      "_________________________________________________________________\n",
      "activation_23 (Activation)   (None, 14, 14, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_23 (MaxPooling (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "dropout_33 (Dropout)         (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_24 (Conv2D)           (None, 5, 5, 128)         295040    \n",
      "_________________________________________________________________\n",
      "activation_24 (Activation)   (None, 5, 5, 128)         0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_24 (MaxPooling (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "dropout_34 (Dropout)         (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten_6 (Flatten)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_16 (Dense)             (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dropout_35 (Dropout)         (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_17 (Dense)             (None, 256)               131328    \n",
      "_________________________________________________________________\n",
      "dropout_36 (Dropout)         (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_18 (Dense)             (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 1,584,257\n",
      "Trainable params: 1,584,257\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\yltbe\\anaconda3\\envs\\tensorflow_gpu\\lib\\site-packages\\ipykernel_launcher.py:11: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  # This is added back by InteractiveShellApp.init_path()\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "===================FOLD= 2\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_7 (InputLayer)         (None, 75, 75, 3)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_25 (Conv2D)           (None, 71, 71, 128)       9728      \n",
      "_________________________________________________________________\n",
      "activation_25 (Activation)   (None, 71, 71, 128)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_25 (MaxPooling (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "dropout_37 (Dropout)         (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_26 (Conv2D)           (None, 33, 33, 256)       295168    \n",
      "_________________________________________________________________\n",
      "activation_26 (Activation)   (None, 33, 33, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_26 (MaxPooling (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "dropout_38 (Dropout)         (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_27 (Conv2D)           (None, 14, 14, 256)       590080    \n",
      "_________________________________________________________________\n",
      "activation_27 (Activation)   (None, 14, 14, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_27 (MaxPooling (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "dropout_39 (Dropout)         (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_28 (Conv2D)           (None, 5, 5, 128)         295040    \n",
      "_________________________________________________________________\n",
      "activation_28 (Activation)   (None, 5, 5, 128)         0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_28 (MaxPooling (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "dropout_40 (Dropout)         (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten_7 (Flatten)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_19 (Dense)             (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dropout_41 (Dropout)         (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_20 (Dense)             (None, 256)               131328    \n",
      "_________________________________________________________________\n",
      "dropout_42 (Dropout)         (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_21 (Dense)             (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 1,584,257\n",
      "Trainable params: 1,584,257\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "\n",
      "===================FOLD= 3\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_8 (InputLayer)         (None, 75, 75, 3)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_29 (Conv2D)           (None, 71, 71, 128)       9728      \n",
      "_________________________________________________________________\n",
      "activation_29 (Activation)   (None, 71, 71, 128)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_29 (MaxPooling (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "dropout_43 (Dropout)         (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_30 (Conv2D)           (None, 33, 33, 256)       295168    \n",
      "_________________________________________________________________\n",
      "activation_30 (Activation)   (None, 33, 33, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_30 (MaxPooling (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "dropout_44 (Dropout)         (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_31 (Conv2D)           (None, 14, 14, 256)       590080    \n",
      "_________________________________________________________________\n",
      "activation_31 (Activation)   (None, 14, 14, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_31 (MaxPooling (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "dropout_45 (Dropout)         (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_32 (Conv2D)           (None, 5, 5, 128)         295040    \n",
      "_________________________________________________________________\n",
      "activation_32 (Activation)   (None, 5, 5, 128)         0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_32 (MaxPooling (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "dropout_46 (Dropout)         (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten_8 (Flatten)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_22 (Dense)             (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dropout_47 (Dropout)         (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_23 (Dense)             (None, 256)               131328    \n",
      "_________________________________________________________________\n",
      "dropout_48 (Dropout)         (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_24 (Dense)             (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 1,584,257\n",
      "Trainable params: 1,584,257\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "\n",
      "===================FOLD= 4\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_9 (InputLayer)         (None, 75, 75, 3)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_33 (Conv2D)           (None, 71, 71, 128)       9728      \n",
      "_________________________________________________________________\n",
      "activation_33 (Activation)   (None, 71, 71, 128)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_33 (MaxPooling (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "dropout_49 (Dropout)         (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_34 (Conv2D)           (None, 33, 33, 256)       295168    \n",
      "_________________________________________________________________\n",
      "activation_34 (Activation)   (None, 33, 33, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_34 (MaxPooling (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "dropout_50 (Dropout)         (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_35 (Conv2D)           (None, 14, 14, 256)       590080    \n",
      "_________________________________________________________________\n",
      "activation_35 (Activation)   (None, 14, 14, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_35 (MaxPooling (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "dropout_51 (Dropout)         (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_36 (Conv2D)           (None, 5, 5, 128)         295040    \n",
      "_________________________________________________________________\n",
      "activation_36 (Activation)   (None, 5, 5, 128)         0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_36 (MaxPooling (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "dropout_52 (Dropout)         (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten_9 (Flatten)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_25 (Dense)             (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dropout_53 (Dropout)         (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_26 (Dense)             (None, 256)               131328    \n",
      "_________________________________________________________________\n",
      "dropout_54 (Dropout)         (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_27 (Dense)             (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 1,584,257\n",
      "Trainable params: 1,584,257\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "===================FOLD= 5\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_10 (InputLayer)        (None, 75, 75, 3)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_37 (Conv2D)           (None, 71, 71, 128)       9728      \n",
      "_________________________________________________________________\n",
      "activation_37 (Activation)   (None, 71, 71, 128)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_37 (MaxPooling (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "dropout_55 (Dropout)         (None, 35, 35, 128)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_38 (Conv2D)           (None, 33, 33, 256)       295168    \n",
      "_________________________________________________________________\n",
      "activation_38 (Activation)   (None, 33, 33, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_38 (MaxPooling (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "dropout_56 (Dropout)         (None, 16, 16, 256)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_39 (Conv2D)           (None, 14, 14, 256)       590080    \n",
      "_________________________________________________________________\n",
      "activation_39 (Activation)   (None, 14, 14, 256)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_39 (MaxPooling (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "dropout_57 (Dropout)         (None, 7, 7, 256)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_40 (Conv2D)           (None, 5, 5, 128)         295040    \n",
      "_________________________________________________________________\n",
      "activation_40 (Activation)   (None, 5, 5, 128)         0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_40 (MaxPooling (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "dropout_58 (Dropout)         (None, 2, 2, 128)         0         \n",
      "_________________________________________________________________\n",
      "flatten_10 (Flatten)         (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_28 (Dense)             (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dropout_59 (Dropout)         (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_29 (Dense)             (None, 256)               131328    \n",
      "_________________________________________________________________\n",
      "dropout_60 (Dropout)         (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_30 (Dense)             (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 1,584,257\n",
      "Trainable params: 1,584,257\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "test_df = test_scores()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>cv1</th>\n",
       "      <th>cv2</th>\n",
       "      <th>cv3</th>\n",
       "      <th>cv4</th>\n",
       "      <th>cv5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5941774d</td>\n",
       "      <td>5.428893e-02</td>\n",
       "      <td>6.673208e-02</td>\n",
       "      <td>7.094257e-02</td>\n",
       "      <td>3.609517e-03</td>\n",
       "      <td>4.734804e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4023181e</td>\n",
       "      <td>5.245490e-01</td>\n",
       "      <td>4.921556e-01</td>\n",
       "      <td>3.735768e-01</td>\n",
       "      <td>2.362068e-01</td>\n",
       "      <td>3.767341e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>b20200e4</td>\n",
       "      <td>7.215895e-06</td>\n",
       "      <td>5.210542e-08</td>\n",
       "      <td>7.925581e-10</td>\n",
       "      <td>2.667083e-08</td>\n",
       "      <td>4.645425e-06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>e7f018bb</td>\n",
       "      <td>9.979774e-01</td>\n",
       "      <td>9.999882e-01</td>\n",
       "      <td>9.976949e-01</td>\n",
       "      <td>9.999983e-01</td>\n",
       "      <td>9.999951e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4371c8c3</td>\n",
       "      <td>1.210454e-01</td>\n",
       "      <td>9.455159e-04</td>\n",
       "      <td>3.490464e-02</td>\n",
       "      <td>3.091512e-03</td>\n",
       "      <td>4.910483e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>a8d9b1fd</td>\n",
       "      <td>8.201791e-02</td>\n",
       "      <td>1.096915e-01</td>\n",
       "      <td>1.754192e-02</td>\n",
       "      <td>6.308465e-02</td>\n",
       "      <td>6.289902e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>29e7727e</td>\n",
       "      <td>2.341684e-01</td>\n",
       "      <td>1.582016e-01</td>\n",
       "      <td>4.415979e-02</td>\n",
       "      <td>3.304548e-01</td>\n",
       "      <td>1.907021e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>92a51ffb</td>\n",
       "      <td>9.991114e-01</td>\n",
       "      <td>9.999897e-01</td>\n",
       "      <td>9.985918e-01</td>\n",
       "      <td>9.999380e-01</td>\n",
       "      <td>9.999481e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>c769ac97</td>\n",
       "      <td>3.057454e-05</td>\n",
       "      <td>1.249167e-04</td>\n",
       "      <td>1.259599e-06</td>\n",
       "      <td>1.479310e-07</td>\n",
       "      <td>2.700044e-09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>aee0547d</td>\n",
       "      <td>1.977355e-09</td>\n",
       "      <td>3.014829e-10</td>\n",
       "      <td>8.679121e-09</td>\n",
       "      <td>9.400190e-12</td>\n",
       "      <td>3.189674e-13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>565b28ac</td>\n",
       "      <td>3.490660e-06</td>\n",
       "      <td>2.864628e-06</td>\n",
       "      <td>1.977537e-06</td>\n",
       "      <td>6.593015e-07</td>\n",
       "      <td>1.076952e-11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>e04e9775</td>\n",
       "      <td>3.843006e-01</td>\n",
       "      <td>1.698858e-01</td>\n",
       "      <td>3.297559e-01</td>\n",
       "      <td>8.405116e-02</td>\n",
       "      <td>4.260120e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>8e8161d1</td>\n",
       "      <td>3.570523e-03</td>\n",
       "      <td>5.131339e-04</td>\n",
       "      <td>3.002876e-03</td>\n",
       "      <td>2.345739e-03</td>\n",
       "      <td>1.461222e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>4cf4d256</td>\n",
       "      <td>5.245575e-01</td>\n",
       "      <td>3.889462e-01</td>\n",
       "      <td>4.365207e-01</td>\n",
       "      <td>2.106485e-01</td>\n",
       "      <td>4.239623e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>139e5324</td>\n",
       "      <td>1.536431e-02</td>\n",
       "      <td>9.576926e-04</td>\n",
       "      <td>2.177959e-03</td>\n",
       "      <td>2.080786e-05</td>\n",
       "      <td>1.467442e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>f156976f</td>\n",
       "      <td>2.907326e-03</td>\n",
       "      <td>4.042191e-03</td>\n",
       "      <td>4.710612e-03</td>\n",
       "      <td>1.065615e-02</td>\n",
       "      <td>3.732556e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>68a117cc</td>\n",
       "      <td>1.454708e-02</td>\n",
       "      <td>4.569225e-04</td>\n",
       "      <td>4.612402e-03</td>\n",
       "      <td>1.036679e-03</td>\n",
       "      <td>4.755504e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>d9aa7a56</td>\n",
       "      <td>1.911590e-02</td>\n",
       "      <td>8.816103e-03</td>\n",
       "      <td>3.873792e-03</td>\n",
       "      <td>6.697680e-03</td>\n",
       "      <td>3.636210e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>9005b143</td>\n",
       "      <td>3.402909e-01</td>\n",
       "      <td>2.137174e-01</td>\n",
       "      <td>2.491119e-02</td>\n",
       "      <td>5.496998e-02</td>\n",
       "      <td>2.578304e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>5f6d3988</td>\n",
       "      <td>9.965498e-01</td>\n",
       "      <td>9.891254e-01</td>\n",
       "      <td>9.423512e-01</td>\n",
       "      <td>9.877314e-01</td>\n",
       "      <td>9.954588e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>9ad70954</td>\n",
       "      <td>2.400184e-02</td>\n",
       "      <td>1.053493e-03</td>\n",
       "      <td>5.176716e-03</td>\n",
       "      <td>8.561841e-04</td>\n",
       "      <td>9.323658e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>b9087b9e</td>\n",
       "      <td>6.003870e-01</td>\n",
       "      <td>5.479059e-01</td>\n",
       "      <td>4.025350e-01</td>\n",
       "      <td>3.204973e-01</td>\n",
       "      <td>9.383065e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>a39a1427</td>\n",
       "      <td>2.704988e-01</td>\n",
       "      <td>1.784840e-01</td>\n",
       "      <td>1.131404e-01</td>\n",
       "      <td>7.899514e-02</td>\n",
       "      <td>2.854400e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>82fbe8ed</td>\n",
       "      <td>1.245286e-01</td>\n",
       "      <td>5.754833e-02</td>\n",
       "      <td>1.732828e-01</td>\n",
       "      <td>5.249222e-02</td>\n",
       "      <td>2.408999e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1fae4879</td>\n",
       "      <td>1.058873e-11</td>\n",
       "      <td>1.017950e-16</td>\n",
       "      <td>2.352451e-10</td>\n",
       "      <td>6.618841e-13</td>\n",
       "      <td>1.065264e-20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>6dd8f13d</td>\n",
       "      <td>6.672931e-05</td>\n",
       "      <td>7.148489e-09</td>\n",
       "      <td>1.598836e-09</td>\n",
       "      <td>2.100684e-07</td>\n",
       "      <td>6.701414e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>bbad5958</td>\n",
       "      <td>7.863383e-02</td>\n",
       "      <td>7.325834e-02</td>\n",
       "      <td>1.093198e-01</td>\n",
       "      <td>2.874385e-02</td>\n",
       "      <td>1.348655e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>54527583</td>\n",
       "      <td>2.228054e-02</td>\n",
       "      <td>3.507486e-03</td>\n",
       "      <td>1.448483e-02</td>\n",
       "      <td>6.297823e-03</td>\n",
       "      <td>2.310304e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>be8fa29c</td>\n",
       "      <td>3.033737e-02</td>\n",
       "      <td>1.204512e-03</td>\n",
       "      <td>1.914037e-02</td>\n",
       "      <td>1.343855e-03</td>\n",
       "      <td>1.088336e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>81a3328f</td>\n",
       "      <td>9.998591e-01</td>\n",
       "      <td>9.999694e-01</td>\n",
       "      <td>9.994621e-01</td>\n",
       "      <td>9.999989e-01</td>\n",
       "      <td>9.999098e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8394</th>\n",
       "      <td>8ae30ce6</td>\n",
       "      <td>2.885115e-04</td>\n",
       "      <td>1.022979e-05</td>\n",
       "      <td>8.558678e-06</td>\n",
       "      <td>6.967151e-06</td>\n",
       "      <td>7.008944e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8395</th>\n",
       "      <td>de27ed88</td>\n",
       "      <td>4.047099e-08</td>\n",
       "      <td>1.652992e-10</td>\n",
       "      <td>2.685429e-06</td>\n",
       "      <td>6.500398e-09</td>\n",
       "      <td>2.105800e-12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8396</th>\n",
       "      <td>66d5196f</td>\n",
       "      <td>1.365313e-01</td>\n",
       "      <td>3.645140e-02</td>\n",
       "      <td>8.939961e-03</td>\n",
       "      <td>3.883383e-02</td>\n",
       "      <td>7.236260e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8397</th>\n",
       "      <td>d85f1858</td>\n",
       "      <td>5.446094e-01</td>\n",
       "      <td>4.933017e-01</td>\n",
       "      <td>4.503037e-01</td>\n",
       "      <td>3.269328e-01</td>\n",
       "      <td>3.972017e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8398</th>\n",
       "      <td>16dcb33a</td>\n",
       "      <td>5.625611e-01</td>\n",
       "      <td>1.971873e-01</td>\n",
       "      <td>3.734035e-01</td>\n",
       "      <td>1.113753e-01</td>\n",
       "      <td>2.227086e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8399</th>\n",
       "      <td>eca3158e</td>\n",
       "      <td>1.141631e-04</td>\n",
       "      <td>3.264449e-07</td>\n",
       "      <td>1.838077e-05</td>\n",
       "      <td>1.914145e-07</td>\n",
       "      <td>6.358623e-11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8400</th>\n",
       "      <td>08daeee6</td>\n",
       "      <td>1.842485e-03</td>\n",
       "      <td>4.092801e-05</td>\n",
       "      <td>3.556691e-04</td>\n",
       "      <td>1.595997e-05</td>\n",
       "      <td>2.471751e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8401</th>\n",
       "      <td>e9c513ee</td>\n",
       "      <td>3.147907e-04</td>\n",
       "      <td>1.490709e-03</td>\n",
       "      <td>4.332528e-05</td>\n",
       "      <td>1.270738e-04</td>\n",
       "      <td>1.490626e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8402</th>\n",
       "      <td>b1519fa6</td>\n",
       "      <td>4.820322e-01</td>\n",
       "      <td>3.819768e-01</td>\n",
       "      <td>4.219830e-01</td>\n",
       "      <td>2.231206e-01</td>\n",
       "      <td>2.721750e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8403</th>\n",
       "      <td>dfc89540</td>\n",
       "      <td>3.216478e-01</td>\n",
       "      <td>6.235651e-02</td>\n",
       "      <td>3.412594e-01</td>\n",
       "      <td>1.321200e-01</td>\n",
       "      <td>2.056761e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8404</th>\n",
       "      <td>8fd8c0e9</td>\n",
       "      <td>5.921112e-01</td>\n",
       "      <td>5.794204e-01</td>\n",
       "      <td>3.941731e-01</td>\n",
       "      <td>3.777530e-01</td>\n",
       "      <td>4.375779e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8405</th>\n",
       "      <td>45df6347</td>\n",
       "      <td>2.718644e-04</td>\n",
       "      <td>2.837665e-05</td>\n",
       "      <td>2.187625e-04</td>\n",
       "      <td>5.412338e-05</td>\n",
       "      <td>8.184473e-06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8406</th>\n",
       "      <td>bf7928d7</td>\n",
       "      <td>6.990940e-04</td>\n",
       "      <td>1.815950e-04</td>\n",
       "      <td>7.309984e-07</td>\n",
       "      <td>4.427364e-05</td>\n",
       "      <td>5.019085e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8407</th>\n",
       "      <td>7b587c05</td>\n",
       "      <td>3.705973e-03</td>\n",
       "      <td>1.394434e-03</td>\n",
       "      <td>4.144018e-03</td>\n",
       "      <td>4.328324e-04</td>\n",
       "      <td>1.035350e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8408</th>\n",
       "      <td>c2834388</td>\n",
       "      <td>2.386024e-01</td>\n",
       "      <td>2.096081e-01</td>\n",
       "      <td>1.736936e-01</td>\n",
       "      <td>2.825594e-02</td>\n",
       "      <td>8.077905e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8409</th>\n",
       "      <td>146143c3</td>\n",
       "      <td>9.999019e-01</td>\n",
       "      <td>9.999996e-01</td>\n",
       "      <td>9.997702e-01</td>\n",
       "      <td>9.999866e-01</td>\n",
       "      <td>9.999996e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8410</th>\n",
       "      <td>d59aee00</td>\n",
       "      <td>6.321745e-09</td>\n",
       "      <td>3.459580e-13</td>\n",
       "      <td>9.390921e-09</td>\n",
       "      <td>1.609377e-11</td>\n",
       "      <td>7.965661e-19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8411</th>\n",
       "      <td>cbc0b93b</td>\n",
       "      <td>7.197064e-01</td>\n",
       "      <td>9.160747e-02</td>\n",
       "      <td>1.500871e-01</td>\n",
       "      <td>5.108611e-02</td>\n",
       "      <td>1.521135e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8412</th>\n",
       "      <td>088e2ff7</td>\n",
       "      <td>9.840918e-01</td>\n",
       "      <td>9.498385e-01</td>\n",
       "      <td>8.269121e-01</td>\n",
       "      <td>8.523738e-01</td>\n",
       "      <td>9.020713e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8413</th>\n",
       "      <td>673d33cd</td>\n",
       "      <td>1.568345e-01</td>\n",
       "      <td>9.584773e-02</td>\n",
       "      <td>2.032875e-02</td>\n",
       "      <td>4.710763e-02</td>\n",
       "      <td>2.737421e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8414</th>\n",
       "      <td>674b031e</td>\n",
       "      <td>1.506795e-05</td>\n",
       "      <td>1.211709e-05</td>\n",
       "      <td>8.925614e-07</td>\n",
       "      <td>3.924320e-03</td>\n",
       "      <td>6.580761e-06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8415</th>\n",
       "      <td>43db4207</td>\n",
       "      <td>3.330339e-01</td>\n",
       "      <td>6.896620e-02</td>\n",
       "      <td>1.355650e-01</td>\n",
       "      <td>1.038796e-01</td>\n",
       "      <td>1.012665e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8416</th>\n",
       "      <td>156855e1</td>\n",
       "      <td>1.458744e-13</td>\n",
       "      <td>7.701359e-23</td>\n",
       "      <td>2.478490e-12</td>\n",
       "      <td>1.003473e-16</td>\n",
       "      <td>3.743123e-36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8417</th>\n",
       "      <td>ac96cfb0</td>\n",
       "      <td>1.450476e-01</td>\n",
       "      <td>1.600608e-02</td>\n",
       "      <td>1.229323e-01</td>\n",
       "      <td>1.191310e-02</td>\n",
       "      <td>3.910888e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8418</th>\n",
       "      <td>fe45aef5</td>\n",
       "      <td>9.905440e-01</td>\n",
       "      <td>9.786154e-01</td>\n",
       "      <td>9.335779e-01</td>\n",
       "      <td>8.809758e-01</td>\n",
       "      <td>7.412164e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8419</th>\n",
       "      <td>16ee9b50</td>\n",
       "      <td>1.570981e-07</td>\n",
       "      <td>4.946981e-07</td>\n",
       "      <td>4.826791e-09</td>\n",
       "      <td>7.817603e-10</td>\n",
       "      <td>4.781171e-12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8420</th>\n",
       "      <td>5a599eb7</td>\n",
       "      <td>3.655356e-01</td>\n",
       "      <td>1.643127e-01</td>\n",
       "      <td>2.257679e-01</td>\n",
       "      <td>1.305295e-01</td>\n",
       "      <td>3.865938e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8421</th>\n",
       "      <td>df30d6dd</td>\n",
       "      <td>8.949722e-03</td>\n",
       "      <td>9.915116e-04</td>\n",
       "      <td>7.453144e-03</td>\n",
       "      <td>2.473858e-03</td>\n",
       "      <td>2.759178e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8422</th>\n",
       "      <td>18af95b1</td>\n",
       "      <td>9.926225e-01</td>\n",
       "      <td>9.959307e-01</td>\n",
       "      <td>9.978351e-01</td>\n",
       "      <td>9.942101e-01</td>\n",
       "      <td>9.999415e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8423</th>\n",
       "      <td>27d788c8</td>\n",
       "      <td>5.997514e-01</td>\n",
       "      <td>8.806776e-01</td>\n",
       "      <td>1.328229e-02</td>\n",
       "      <td>3.145537e-02</td>\n",
       "      <td>8.568646e-01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8424 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            id           cv1           cv2           cv3           cv4  \\\n",
       "0     5941774d  5.428893e-02  6.673208e-02  7.094257e-02  3.609517e-03   \n",
       "1     4023181e  5.245490e-01  4.921556e-01  3.735768e-01  2.362068e-01   \n",
       "2     b20200e4  7.215895e-06  5.210542e-08  7.925581e-10  2.667083e-08   \n",
       "3     e7f018bb  9.979774e-01  9.999882e-01  9.976949e-01  9.999983e-01   \n",
       "4     4371c8c3  1.210454e-01  9.455159e-04  3.490464e-02  3.091512e-03   \n",
       "5     a8d9b1fd  8.201791e-02  1.096915e-01  1.754192e-02  6.308465e-02   \n",
       "6     29e7727e  2.341684e-01  1.582016e-01  4.415979e-02  3.304548e-01   \n",
       "7     92a51ffb  9.991114e-01  9.999897e-01  9.985918e-01  9.999380e-01   \n",
       "8     c769ac97  3.057454e-05  1.249167e-04  1.259599e-06  1.479310e-07   \n",
       "9     aee0547d  1.977355e-09  3.014829e-10  8.679121e-09  9.400190e-12   \n",
       "10    565b28ac  3.490660e-06  2.864628e-06  1.977537e-06  6.593015e-07   \n",
       "11    e04e9775  3.843006e-01  1.698858e-01  3.297559e-01  8.405116e-02   \n",
       "12    8e8161d1  3.570523e-03  5.131339e-04  3.002876e-03  2.345739e-03   \n",
       "13    4cf4d256  5.245575e-01  3.889462e-01  4.365207e-01  2.106485e-01   \n",
       "14    139e5324  1.536431e-02  9.576926e-04  2.177959e-03  2.080786e-05   \n",
       "15    f156976f  2.907326e-03  4.042191e-03  4.710612e-03  1.065615e-02   \n",
       "16    68a117cc  1.454708e-02  4.569225e-04  4.612402e-03  1.036679e-03   \n",
       "17    d9aa7a56  1.911590e-02  8.816103e-03  3.873792e-03  6.697680e-03   \n",
       "18    9005b143  3.402909e-01  2.137174e-01  2.491119e-02  5.496998e-02   \n",
       "19    5f6d3988  9.965498e-01  9.891254e-01  9.423512e-01  9.877314e-01   \n",
       "20    9ad70954  2.400184e-02  1.053493e-03  5.176716e-03  8.561841e-04   \n",
       "21    b9087b9e  6.003870e-01  5.479059e-01  4.025350e-01  3.204973e-01   \n",
       "22    a39a1427  2.704988e-01  1.784840e-01  1.131404e-01  7.899514e-02   \n",
       "23    82fbe8ed  1.245286e-01  5.754833e-02  1.732828e-01  5.249222e-02   \n",
       "24    1fae4879  1.058873e-11  1.017950e-16  2.352451e-10  6.618841e-13   \n",
       "25    6dd8f13d  6.672931e-05  7.148489e-09  1.598836e-09  2.100684e-07   \n",
       "26    bbad5958  7.863383e-02  7.325834e-02  1.093198e-01  2.874385e-02   \n",
       "27    54527583  2.228054e-02  3.507486e-03  1.448483e-02  6.297823e-03   \n",
       "28    be8fa29c  3.033737e-02  1.204512e-03  1.914037e-02  1.343855e-03   \n",
       "29    81a3328f  9.998591e-01  9.999694e-01  9.994621e-01  9.999989e-01   \n",
       "...        ...           ...           ...           ...           ...   \n",
       "8394  8ae30ce6  2.885115e-04  1.022979e-05  8.558678e-06  6.967151e-06   \n",
       "8395  de27ed88  4.047099e-08  1.652992e-10  2.685429e-06  6.500398e-09   \n",
       "8396  66d5196f  1.365313e-01  3.645140e-02  8.939961e-03  3.883383e-02   \n",
       "8397  d85f1858  5.446094e-01  4.933017e-01  4.503037e-01  3.269328e-01   \n",
       "8398  16dcb33a  5.625611e-01  1.971873e-01  3.734035e-01  1.113753e-01   \n",
       "8399  eca3158e  1.141631e-04  3.264449e-07  1.838077e-05  1.914145e-07   \n",
       "8400  08daeee6  1.842485e-03  4.092801e-05  3.556691e-04  1.595997e-05   \n",
       "8401  e9c513ee  3.147907e-04  1.490709e-03  4.332528e-05  1.270738e-04   \n",
       "8402  b1519fa6  4.820322e-01  3.819768e-01  4.219830e-01  2.231206e-01   \n",
       "8403  dfc89540  3.216478e-01  6.235651e-02  3.412594e-01  1.321200e-01   \n",
       "8404  8fd8c0e9  5.921112e-01  5.794204e-01  3.941731e-01  3.777530e-01   \n",
       "8405  45df6347  2.718644e-04  2.837665e-05  2.187625e-04  5.412338e-05   \n",
       "8406  bf7928d7  6.990940e-04  1.815950e-04  7.309984e-07  4.427364e-05   \n",
       "8407  7b587c05  3.705973e-03  1.394434e-03  4.144018e-03  4.328324e-04   \n",
       "8408  c2834388  2.386024e-01  2.096081e-01  1.736936e-01  2.825594e-02   \n",
       "8409  146143c3  9.999019e-01  9.999996e-01  9.997702e-01  9.999866e-01   \n",
       "8410  d59aee00  6.321745e-09  3.459580e-13  9.390921e-09  1.609377e-11   \n",
       "8411  cbc0b93b  7.197064e-01  9.160747e-02  1.500871e-01  5.108611e-02   \n",
       "8412  088e2ff7  9.840918e-01  9.498385e-01  8.269121e-01  8.523738e-01   \n",
       "8413  673d33cd  1.568345e-01  9.584773e-02  2.032875e-02  4.710763e-02   \n",
       "8414  674b031e  1.506795e-05  1.211709e-05  8.925614e-07  3.924320e-03   \n",
       "8415  43db4207  3.330339e-01  6.896620e-02  1.355650e-01  1.038796e-01   \n",
       "8416  156855e1  1.458744e-13  7.701359e-23  2.478490e-12  1.003473e-16   \n",
       "8417  ac96cfb0  1.450476e-01  1.600608e-02  1.229323e-01  1.191310e-02   \n",
       "8418  fe45aef5  9.905440e-01  9.786154e-01  9.335779e-01  8.809758e-01   \n",
       "8419  16ee9b50  1.570981e-07  4.946981e-07  4.826791e-09  7.817603e-10   \n",
       "8420  5a599eb7  3.655356e-01  1.643127e-01  2.257679e-01  1.305295e-01   \n",
       "8421  df30d6dd  8.949722e-03  9.915116e-04  7.453144e-03  2.473858e-03   \n",
       "8422  18af95b1  9.926225e-01  9.959307e-01  9.978351e-01  9.942101e-01   \n",
       "8423  27d788c8  5.997514e-01  8.806776e-01  1.328229e-02  3.145537e-02   \n",
       "\n",
       "               cv5  \n",
       "0     4.734804e-02  \n",
       "1     3.767341e-01  \n",
       "2     4.645425e-06  \n",
       "3     9.999951e-01  \n",
       "4     4.910483e-05  \n",
       "5     6.289902e-02  \n",
       "6     1.907021e-01  \n",
       "7     9.999481e-01  \n",
       "8     2.700044e-09  \n",
       "9     3.189674e-13  \n",
       "10    1.076952e-11  \n",
       "11    4.260120e-02  \n",
       "12    1.461222e-04  \n",
       "13    4.239623e-01  \n",
       "14    1.467442e-04  \n",
       "15    3.732556e-04  \n",
       "16    4.755504e-05  \n",
       "17    3.636210e-03  \n",
       "18    2.578304e-01  \n",
       "19    9.954588e-01  \n",
       "20    9.323658e-05  \n",
       "21    9.383065e-02  \n",
       "22    2.854400e-02  \n",
       "23    2.408999e-02  \n",
       "24    1.065264e-20  \n",
       "25    6.701414e-07  \n",
       "26    1.348655e-03  \n",
       "27    2.310304e-04  \n",
       "28    1.088336e-04  \n",
       "29    9.999098e-01  \n",
       "...            ...  \n",
       "8394  7.008944e-07  \n",
       "8395  2.105800e-12  \n",
       "8396  7.236260e-02  \n",
       "8397  3.972017e-01  \n",
       "8398  2.227086e-01  \n",
       "8399  6.358623e-11  \n",
       "8400  2.471751e-05  \n",
       "8401  1.490626e-05  \n",
       "8402  2.721750e-01  \n",
       "8403  2.056761e-02  \n",
       "8404  4.375779e-01  \n",
       "8405  8.184473e-06  \n",
       "8406  5.019085e-07  \n",
       "8407  1.035350e-04  \n",
       "8408  8.077905e-03  \n",
       "8409  9.999996e-01  \n",
       "8410  7.965661e-19  \n",
       "8411  1.521135e-01  \n",
       "8412  9.020713e-01  \n",
       "8413  2.737421e-03  \n",
       "8414  6.580761e-06  \n",
       "8415  1.012665e-02  \n",
       "8416  3.743123e-36  \n",
       "8417  3.910888e-04  \n",
       "8418  7.412164e-01  \n",
       "8419  4.781171e-12  \n",
       "8420  3.865938e-02  \n",
       "8421  2.759178e-05  \n",
       "8422  9.999415e-01  \n",
       "8423  8.568646e-01  \n",
       "\n",
       "[8424 rows x 6 columns]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_sub_mmm(test_df):\n",
    "    prediction = test_df\n",
    "    prediction['is_iceberg'] = np.where(np.all(prediction[prediction.columns[1:]]>0.9,axis = 1),\n",
    "                                    np.max(prediction[prediction.columns[1:]],axis = 1),\n",
    "                                     np.where(np.all(prediction[prediction.columns[1:]]<0.1,axis = 1),np.min(prediction[prediction.columns[1:]],axis = 1),np.median(prediction[prediction.columns[1:]],axis = 1)))\n",
    "    sub = prediction[['id','is_iceberg']]\n",
    "    sub.to_csv('sub.csv',index = False)\n",
    "    return sub"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\yltbe\\anaconda3\\envs\\tensorflow_gpu\\lib\\site-packages\\ipykernel_launcher.py:5: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  \"\"\"\n"
     ]
    }
   ],
   "source": [
    "sub = get_sub_mmm(test_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>is_iceberg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5941774d</td>\n",
       "      <td>3.609517e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4023181e</td>\n",
       "      <td>3.767341e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>b20200e4</td>\n",
       "      <td>7.925581e-10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>e7f018bb</td>\n",
       "      <td>9.999983e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4371c8c3</td>\n",
       "      <td>3.091512e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>a8d9b1fd</td>\n",
       "      <td>6.308465e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>29e7727e</td>\n",
       "      <td>1.907021e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>92a51ffb</td>\n",
       "      <td>9.999897e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>c769ac97</td>\n",
       "      <td>2.700044e-09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>aee0547d</td>\n",
       "      <td>3.189674e-13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>565b28ac</td>\n",
       "      <td>1.076952e-11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>e04e9775</td>\n",
       "      <td>1.698858e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>8e8161d1</td>\n",
       "      <td>1.461222e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>4cf4d256</td>\n",
       "      <td>4.239623e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>139e5324</td>\n",
       "      <td>2.080786e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>f156976f</td>\n",
       "      <td>3.732556e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>68a117cc</td>\n",
       "      <td>4.755504e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>d9aa7a56</td>\n",
       "      <td>3.636210e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>9005b143</td>\n",
       "      <td>2.137174e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>5f6d3988</td>\n",
       "      <td>9.965498e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>9ad70954</td>\n",
       "      <td>9.323658e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>b9087b9e</td>\n",
       "      <td>4.025350e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>a39a1427</td>\n",
       "      <td>1.131404e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>82fbe8ed</td>\n",
       "      <td>5.754833e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1fae4879</td>\n",
       "      <td>1.065264e-20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>6dd8f13d</td>\n",
       "      <td>1.598836e-09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>bbad5958</td>\n",
       "      <td>7.325834e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>54527583</td>\n",
       "      <td>2.310304e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>be8fa29c</td>\n",
       "      <td>1.088336e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>81a3328f</td>\n",
       "      <td>9.999989e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8394</th>\n",
       "      <td>8ae30ce6</td>\n",
       "      <td>7.008944e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8395</th>\n",
       "      <td>de27ed88</td>\n",
       "      <td>2.105800e-12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8396</th>\n",
       "      <td>66d5196f</td>\n",
       "      <td>3.883383e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8397</th>\n",
       "      <td>d85f1858</td>\n",
       "      <td>4.503037e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8398</th>\n",
       "      <td>16dcb33a</td>\n",
       "      <td>2.227086e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8399</th>\n",
       "      <td>eca3158e</td>\n",
       "      <td>6.358623e-11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8400</th>\n",
       "      <td>08daeee6</td>\n",
       "      <td>1.595997e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8401</th>\n",
       "      <td>e9c513ee</td>\n",
       "      <td>1.490626e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8402</th>\n",
       "      <td>b1519fa6</td>\n",
       "      <td>3.819768e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8403</th>\n",
       "      <td>dfc89540</td>\n",
       "      <td>1.321200e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8404</th>\n",
       "      <td>8fd8c0e9</td>\n",
       "      <td>4.375779e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8405</th>\n",
       "      <td>45df6347</td>\n",
       "      <td>8.184473e-06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8406</th>\n",
       "      <td>bf7928d7</td>\n",
       "      <td>5.019085e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8407</th>\n",
       "      <td>7b587c05</td>\n",
       "      <td>1.035350e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8408</th>\n",
       "      <td>c2834388</td>\n",
       "      <td>1.736936e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8409</th>\n",
       "      <td>146143c3</td>\n",
       "      <td>9.999996e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8410</th>\n",
       "      <td>d59aee00</td>\n",
       "      <td>7.965661e-19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8411</th>\n",
       "      <td>cbc0b93b</td>\n",
       "      <td>1.500871e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8412</th>\n",
       "      <td>088e2ff7</td>\n",
       "      <td>9.020713e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8413</th>\n",
       "      <td>673d33cd</td>\n",
       "      <td>4.710763e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8414</th>\n",
       "      <td>674b031e</td>\n",
       "      <td>8.925614e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8415</th>\n",
       "      <td>43db4207</td>\n",
       "      <td>1.038796e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8416</th>\n",
       "      <td>156855e1</td>\n",
       "      <td>3.743123e-36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8417</th>\n",
       "      <td>ac96cfb0</td>\n",
       "      <td>1.600608e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8418</th>\n",
       "      <td>fe45aef5</td>\n",
       "      <td>9.335779e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8419</th>\n",
       "      <td>16ee9b50</td>\n",
       "      <td>4.781171e-12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8420</th>\n",
       "      <td>5a599eb7</td>\n",
       "      <td>1.643127e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8421</th>\n",
       "      <td>df30d6dd</td>\n",
       "      <td>2.759178e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8422</th>\n",
       "      <td>18af95b1</td>\n",
       "      <td>9.999415e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8423</th>\n",
       "      <td>27d788c8</td>\n",
       "      <td>5.997514e-01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8424 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            id    is_iceberg\n",
       "0     5941774d  3.609517e-03\n",
       "1     4023181e  3.767341e-01\n",
       "2     b20200e4  7.925581e-10\n",
       "3     e7f018bb  9.999983e-01\n",
       "4     4371c8c3  3.091512e-03\n",
       "5     a8d9b1fd  6.308465e-02\n",
       "6     29e7727e  1.907021e-01\n",
       "7     92a51ffb  9.999897e-01\n",
       "8     c769ac97  2.700044e-09\n",
       "9     aee0547d  3.189674e-13\n",
       "10    565b28ac  1.076952e-11\n",
       "11    e04e9775  1.698858e-01\n",
       "12    8e8161d1  1.461222e-04\n",
       "13    4cf4d256  4.239623e-01\n",
       "14    139e5324  2.080786e-05\n",
       "15    f156976f  3.732556e-04\n",
       "16    68a117cc  4.755504e-05\n",
       "17    d9aa7a56  3.636210e-03\n",
       "18    9005b143  2.137174e-01\n",
       "19    5f6d3988  9.965498e-01\n",
       "20    9ad70954  9.323658e-05\n",
       "21    b9087b9e  4.025350e-01\n",
       "22    a39a1427  1.131404e-01\n",
       "23    82fbe8ed  5.754833e-02\n",
       "24    1fae4879  1.065264e-20\n",
       "25    6dd8f13d  1.598836e-09\n",
       "26    bbad5958  7.325834e-02\n",
       "27    54527583  2.310304e-04\n",
       "28    be8fa29c  1.088336e-04\n",
       "29    81a3328f  9.999989e-01\n",
       "...        ...           ...\n",
       "8394  8ae30ce6  7.008944e-07\n",
       "8395  de27ed88  2.105800e-12\n",
       "8396  66d5196f  3.883383e-02\n",
       "8397  d85f1858  4.503037e-01\n",
       "8398  16dcb33a  2.227086e-01\n",
       "8399  eca3158e  6.358623e-11\n",
       "8400  08daeee6  1.595997e-05\n",
       "8401  e9c513ee  1.490626e-05\n",
       "8402  b1519fa6  3.819768e-01\n",
       "8403  dfc89540  1.321200e-01\n",
       "8404  8fd8c0e9  4.375779e-01\n",
       "8405  45df6347  8.184473e-06\n",
       "8406  bf7928d7  5.019085e-07\n",
       "8407  7b587c05  1.035350e-04\n",
       "8408  c2834388  1.736936e-01\n",
       "8409  146143c3  9.999996e-01\n",
       "8410  d59aee00  7.965661e-19\n",
       "8411  cbc0b93b  1.500871e-01\n",
       "8412  088e2ff7  9.020713e-01\n",
       "8413  673d33cd  4.710763e-02\n",
       "8414  674b031e  8.925614e-07\n",
       "8415  43db4207  1.038796e-01\n",
       "8416  156855e1  3.743123e-36\n",
       "8417  ac96cfb0  1.600608e-02\n",
       "8418  fe45aef5  9.335779e-01\n",
       "8419  16ee9b50  4.781171e-12\n",
       "8420  5a599eb7  1.643127e-01\n",
       "8421  df30d6dd  2.759178e-05\n",
       "8422  18af95b1  9.999415e-01\n",
       "8423  27d788c8  5.997514e-01\n",
       "\n",
       "[8424 rows x 2 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sub"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}