class_05_add_MNIST.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# MNIST (Simple)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "#-*- coding: utf-8 -*-\n",
    "\n",
    "from keras.datasets import mnist\n",
    "from keras.utils import np_utils\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPooling2D\n",
    "from keras.callbacks import ModelCheckpoint,EarlyStopping\n",
    "import numpy\n",
    "import sys\n",
    "import os\n",
    "import tensorflow as tf\n",
    "\n",
    "# set seed \n",
    "seed = 0\n",
    "numpy.random.seed(seed)\n",
    "tf.random.set_seed(seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# of training images : 60000 \n",
      "# of testing images : 10000 \n"
     ]
    }
   ],
   "source": [
    "# load MNIST Data\n",
    "(X_train, Y_train), (X_test, Y_test) = mnist.load_data()\n",
    "\n",
    "print(\"# of training images : %d \" % (X_train.shape[0]))\n",
    "print(\"# of testing images : %d \" % (X_test.shape[0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 28, 28)\n",
      "(10000, 28, 28)\n",
      "(60000,)\n",
      "(10000,)\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape)\n",
    "print(X_test.shape)\n",
    "print(Y_train.shape)\n",
    "print(Y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000,)"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD4CAYAAAAq5pAIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAOS0lEQVR4nO3df4xU9bnH8c8jgqgQg7JQYsnd3kZNjcnd4kiuQQiXegnyDxDsTUlsaCTdxh9JMcRcszex/kgMMZdWjKbJ9oLQm15rFRBMzC1KSAyJVkdFBfF31rIFYYlKhSgt8Nw/9nCz4sx3lpkzc4Z93q9kMzPnOWfP47gfzsx8z5mvubsAjHznFN0AgNYg7EAQhB0IgrADQRB2IIhzW7mziRMnemdnZyt3CYTS19enQ4cOWaVaQ2E3s3mSVksaJem/3H1lav3Ozk6Vy+VGdgkgoVQqVa3V/TLezEZJelTSDZKulLTEzK6s9/cBaK5G3rNPl/SBu3/k7n+T9HtJC/JpC0DeGgn7pZL2Dnncny37GjPrNrOymZUHBgYa2B2ARjQS9kofAnzj3Ft373X3kruXOjo6GtgdgEY0EvZ+SVOHPP62pH2NtQOgWRoJ+yuSLjOz75jZGEk/krQln7YA5K3uoTd3P25mt0v6owaH3ta6++7cOgOQq4bG2d39WUnP5tQLgCbidFkgCMIOBEHYgSAIOxAEYQeCIOxAEIQdCIKwA0EQdiAIwg4EQdiBIAg7EARhB4Ig7EAQhB0IgrADQRB2IAjCDgRB2IEgCDsQBGEHgiDsQBCEHQiCsANBEHYgCMIOBEHYgSAIOxAEYQeCaGgWV7S/kydPJuvHjh1r6v7Xr19ftXb06NHktm+//Xay/tBDDyXrPT09VWuPPPJIctvzzz8/WV+1alWyfssttyTrRWgo7GbWJ+kLSSckHXf3Uh5NAchfHkf2f3H3Qzn8HgBNxHt2IIhGw+6StprZq2bWXWkFM+s2s7KZlQcGBhrcHYB6NRr2Ge4+TdINkm4zs1mnr+Duve5ecvdSR0dHg7sDUK+Gwu7u+7Lbg5I2SZqeR1MA8ld32M3sQjMbf+q+pLmSduXVGIB8NfJp/GRJm8zs1O/5H3f/31y6GmEOHz6crJ84cSJZf+ONN5L1rVu3Vq19/vnnyW17e3uT9SJ1dnYm6ytWrEjW16xZU7V20UUXJbedOXNmsj5nzpxkvR3VHXZ3/0jSP+XYC4AmYugNCIKwA0EQdiAIwg4EQdiBILjENQf9/f3JeldXV7L+2Wef5dnOWeOcc9LHmtTQmVT7MtRly5ZVrU2aNCm57bhx45L1s/FsUI7sQBCEHQiCsANBEHYgCMIOBEHYgSAIOxAE4+w5uOSSS5L1yZMnJ+vtPM4+d+7cZL3Wf/vGjRur1s4777zktrNnz07WcWY4sgNBEHYgCMIOBEHYgSAIOxAEYQeCIOxAEIyz56DWddXr1q1L1p966qlk/dprr03WFy9enKynXHfddcn65s2bk/UxY8Yk65988knV2urVq5PbIl8c2YEgCDsQBGEHgiDsQBCEHQiCsANBEHYgCHP3lu2sVCp5uVxu2f7OFseOHUvWa41l9/T0VK09+OCDyW23b9+erM+aNStZR3splUoql8tWqVbzyG5ma83soJntGrLsYjN7zszez24n5NkwgPwN52X8OknzTlt2l6Rt7n6ZpG3ZYwBtrGbY3f0FSZ+etniBpPXZ/fWSFubcF4Cc1fsB3WR33y9J2W3VibPMrNvMymZWHhgYqHN3ABrV9E/j3b3X3UvuXjobJ8MDRop6w37AzKZIUnZ7ML+WADRDvWHfImlpdn+ppPR1kAAKV/N6djN7XNJsSRPNrF/SLyStlPQHM1sm6c+SftjMJke6Wt+fXsuECfWPfD788MPJ+syZM5N1s4pDumhDNcPu7kuqlH6Qcy8AmojTZYEgCDsQBGEHgiDsQBCEHQiCr5IeAZYvX1619vLLLye33bRpU7K+e/fuZP2qq65K1tE+OLIDQRB2IAjCDgRB2IEgCDsQBGEHgiDsQBCMs48Aqa+a7u3tTW67bdu2ZH3BggXJ+sKF6a8fnDFjRtXaokWLktty+Wy+OLIDQRB2IAjCDgRB2IEgCDsQBGEHgiDsQBBM2Rxcrevd5807fU7Przt8+HDd+167dm2yvnjx4mR93Lhxde97pGpoymYAIwNhB4Ig7EAQhB0IgrADQRB2IAjCDgTB9ezBTZ8+PVmv9b3xd9xxR7L+5JNPVq3dfPPNyW0//PDDZP3OO+9M1sePH5+sR1PzyG5ma83soJntGrLsHjP7i5ntzH7mN7dNAI0azsv4dZIqnUb1K3fvyn6ezbctAHmrGXZ3f0HSpy3oBUATNfIB3e1m9mb2Mn9CtZXMrNvMymZWHhgYaGB3ABpRb9h/Lem7krok7Ze0qtqK7t7r7iV3L3V0dNS5OwCNqivs7n7A3U+4+0lJv5GU/kgXQOHqCruZTRnycJGkXdXWBdAeal7PbmaPS5otaaKkA5J+kT3ukuSS+iT9zN3319oZ17OPPF999VWy/tJLL1WtXX/99clta/1t3njjjcn6E088kayPRKnr2WueVOPuSyosXtNwVwBaitNlgSAIOxAEYQeCIOxAEIQdCIJLXNGQsWPHJuuzZ8+uWhs1alRy2+PHjyfrTz/9dLL+7rvvVq1dccUVyW1HIo7sQBCEHQiCsANBEHYgCMIOBEHYgSAIOxAE4+xI2rdvX7K+cePGZP3FF1+sWqs1jl7LNddck6xffvnlDf3+kYYjOxAEYQeCIOxAEIQdCIKwA0EQdiAIwg4EwTj7CFdryq1HH300WX/ssceS9f7+/jPuabhqXe/e2dmZrJtV/EblsDiyA0EQdiAIwg4EQdiBIAg7EARhB4Ig7EAQjLOfBY4cOZKsP/PMM1Vr9913X3Lb9957r66e8jBnzpxkfeXKlcn61VdfnWc7I17NI7uZTTWz7Wa2x8x2m9nPs+UXm9lzZvZ+djuh+e0CqNdwXsYfl7TC3b8n6Z8l3WZmV0q6S9I2d79M0rbsMYA2VTPs7r7f3V/L7n8haY+kSyUtkLQ+W229pIXNahJA487oAzoz65T0fUl/kjTZ3fdLg/8gSJpUZZtuMyubWbnWedoAmmfYYTezcZI2SFru7n8d7nbu3uvuJXcvdXR01NMjgBwMK+xmNlqDQf+du5/6OtEDZjYlq0+RdLA5LQLIQ82hNxu8TnCNpD3u/sshpS2Slkpamd1ubkqHI8DRo0eT9b179ybrN910U7L++uuvn3FPeZk7d26yfu+991at1foqaC5RzddwxtlnSPqxpLfMbGe2rEeDIf+DmS2T9GdJP2xOiwDyUDPs7r5DUrV/Yn+QbzsAmoXTZYEgCDsQBGEHgiDsQBCEHQiCS1yH6csvv6xaW758eXLbHTt2JOvvvPNOXT3lYf78+cn63Xffnax3dXUl66NHjz7jntAcHNmBIAg7EARhB4Ig7EAQhB0IgrADQRB2IIgw4+x9fX3J+gMPPJCsP//881VrH3/8cT0t5eaCCy6oWrv//vuT2956663J+pgxY+rqCe2HIzsQBGEHgiDsQBCEHQiCsANBEHYgCMIOBBFmnH3Dhg3J+po1a5q272nTpiXrS5YsSdbPPTf9v6m7u7tqbezYscltEQdHdiAIwg4EQdiBIAg7EARhB4Ig7EAQhB0Iwtw9vYLZVEm/lfQtSScl9br7ajO7R9JPJQ1kq/a4+7Op31UqlbxcLjfcNIDKSqWSyuVyxVmXh3NSzXFJK9z9NTMbL+lVM3suq/3K3f8zr0YBNM9w5mffL2l/dv8LM9sj6dJmNwYgX2f0nt3MOiV9X9KfskW3m9mbZrbWzCZU2abbzMpmVh4YGKi0CoAWGHbYzWycpA2Slrv7XyX9WtJ3JXVp8Mi/qtJ27t7r7iV3L3V0dOTQMoB6DCvsZjZag0H/nbtvlCR3P+DuJ9z9pKTfSJrevDYBNKpm2M3MJK2RtMfdfzlk+ZQhqy2StCv/9gDkZTifxs+Q9GNJb5nZzmxZj6QlZtYlySX1SfpZUzoEkIvhfBq/Q1KlcbvkmDqA9sIZdEAQhB0IgrADQRB2IAjCDgRB2IEgCDsQBGEHgiDsQBCEHQiCsANBEHYgCMIOBEHYgSBqfpV0rjszG5D08ZBFEyUdalkDZ6Zde2vXviR6q1eevf2Du1f8/reWhv0bOzcru3upsAYS2rW3du1Lord6tao3XsYDQRB2IIiiw95b8P5T2rW3du1Lord6taS3Qt+zA2idoo/sAFqEsANBFBJ2M5tnZu+a2QdmdlcRPVRjZn1m9paZ7TSzQueXzubQO2hmu4Ysu9jMnjOz97PbinPsFdTbPWb2l+y522lm8wvqbaqZbTezPWa228x+ni0v9LlL9NWS563l79nNbJSk9yT9q6R+Sa9IWuLub7e0kSrMrE9Syd0LPwHDzGZJOiLpt+5+VbbsQUmfuvvK7B/KCe7+723S2z2SjhQ9jXc2W9GUodOMS1oo6Scq8LlL9PVvasHzVsSRfbqkD9z9I3f/m6TfS1pQQB9tz91fkPTpaYsXSFqf3V+vwT+WlqvSW1tw9/3u/lp2/wtJp6YZL/S5S/TVEkWE/VJJe4c87ld7zffukraa2atm1l10MxVMdvf90uAfj6RJBfdzuprTeLfSadOMt81zV8/0540qIuyVppJqp/G/Ge4+TdINkm7LXq5ieIY1jXerVJhmvC3UO/15o4oIe7+kqUMef1vSvgL6qMjd92W3ByVtUvtNRX3g1Ay62e3Bgvv5f+00jXelacbVBs9dkdOfFxH2VyRdZmbfMbMxkn4kaUsBfXyDmV2YfXAiM7tQ0ly131TUWyQtze4vlbS5wF6+pl2m8a42zbgKfu4Kn/7c3Vv+I2m+Bj+R/1DSfxTRQ5W+/lHSG9nP7qJ7k/S4Bl/W/V2Dr4iWSbpE0jZJ72e3F7dRb/8t6S1Jb2owWFMK6u06Db41fFPSzuxnftHPXaKvljxvnC4LBMEZdEAQhB0IgrADQRB2IAjCDgRB2IEgCDsQxP8BwfxNbNfq1cUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visual check\n",
    "import matplotlib.pyplot as plt\n",
    "plt.imshow(X_train[0], cmap='Greys')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t3\t18\t18\t18\t126\t136\t175\t26\t166\t255\t247\t127\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t30\t36\t94\t154\t170\t253\t253\t253\t253\t253\t225\t172\t253\t242\t195\t64\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t49\t238\t253\t253\t253\t253\t253\t253\t253\t253\t251\t93\t82\t82\t56\t39\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t18\t219\t253\t253\t253\t253\t253\t198\t182\t247\t241\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t80\t156\t107\t253\t253\t205\t11\t0\t43\t154\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t14\t1\t154\t253\t90\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t139\t253\t190\t2\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t11\t190\t253\t70\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t35\t241\t225\t160\t108\t1\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t81\t240\t253\t253\t119\t25\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t45\t186\t253\t253\t150\t27\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t16\t93\t252\t253\t187\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t249\t253\t249\t64\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t46\t130\t183\t253\t253\t207\t2\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t39\t148\t229\t253\t253\t253\t250\t182\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t24\t114\t221\t253\t253\t253\t253\t201\t78\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t23\t66\t213\t253\t253\t253\t253\t198\t81\t2\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t18\t171\t219\t253\t253\t253\t253\t195\t80\t9\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t55\t172\t226\t253\t253\t253\t253\t244\t133\t11\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t136\t253\t253\t253\t212\t135\t132\t16\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n",
      "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t0\t\n"
     ]
    }
   ],
   "source": [
    "# Code No Check\n",
    "for x in X_train[0]:\n",
    "    for i in x:\n",
    "        sys.stdout.write('%d\\t' % i)\n",
    "    sys.stdout.write('\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Reshape\n",
    "X_train = X_train.reshape(X_train.shape[0], 784)\n",
    "X_train = X_train.astype('float64')\n",
    "X_train = X_train / 255\n",
    "\n",
    "X_test = X_test.reshape(X_test.shape[0], 784).astype('float64') / 255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.01176471 0.07058824 0.07058824 0.07058824\n",
      " 0.49411765 0.53333333 0.68627451 0.10196078 0.65098039 1.\n",
      " 0.96862745 0.49803922 0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.11764706 0.14117647 0.36862745 0.60392157\n",
      " 0.66666667 0.99215686 0.99215686 0.99215686 0.99215686 0.99215686\n",
      " 0.88235294 0.6745098  0.99215686 0.94901961 0.76470588 0.25098039\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.19215686\n",
      " 0.93333333 0.99215686 0.99215686 0.99215686 0.99215686 0.99215686\n",
      " 0.99215686 0.99215686 0.99215686 0.98431373 0.36470588 0.32156863\n",
      " 0.32156863 0.21960784 0.15294118 0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.07058824 0.85882353 0.99215686\n",
      " 0.99215686 0.99215686 0.99215686 0.99215686 0.77647059 0.71372549\n",
      " 0.96862745 0.94509804 0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.31372549 0.61176471 0.41960784 0.99215686\n",
      " 0.99215686 0.80392157 0.04313725 0.         0.16862745 0.60392157\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.05490196 0.00392157 0.60392157 0.99215686 0.35294118\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.54509804 0.99215686 0.74509804 0.00784314 0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.04313725\n",
      " 0.74509804 0.99215686 0.2745098  0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.1372549  0.94509804\n",
      " 0.88235294 0.62745098 0.42352941 0.00392157 0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.31764706 0.94117647 0.99215686\n",
      " 0.99215686 0.46666667 0.09803922 0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.17647059 0.72941176 0.99215686 0.99215686\n",
      " 0.58823529 0.10588235 0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.0627451  0.36470588 0.98823529 0.99215686 0.73333333\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.97647059 0.99215686 0.97647059 0.25098039 0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.18039216 0.50980392 0.71764706 0.99215686\n",
      " 0.99215686 0.81176471 0.00784314 0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.15294118 0.58039216\n",
      " 0.89803922 0.99215686 0.99215686 0.99215686 0.98039216 0.71372549\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.09411765 0.44705882 0.86666667 0.99215686 0.99215686 0.99215686\n",
      " 0.99215686 0.78823529 0.30588235 0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.09019608 0.25882353 0.83529412 0.99215686\n",
      " 0.99215686 0.99215686 0.99215686 0.77647059 0.31764706 0.00784314\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.07058824 0.67058824\n",
      " 0.85882353 0.99215686 0.99215686 0.99215686 0.99215686 0.76470588\n",
      " 0.31372549 0.03529412 0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.21568627 0.6745098  0.88627451 0.99215686 0.99215686 0.99215686\n",
      " 0.99215686 0.95686275 0.52156863 0.04313725 0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.53333333 0.99215686\n",
      " 0.99215686 0.99215686 0.83137255 0.52941176 0.51764706 0.0627451\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.         0.         0.\n",
      " 0.         0.         0.         0.        ]\n"
     ]
    }
   ],
   "source": [
    "print(X_train[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "class : 5 \n"
     ]
    }
   ],
   "source": [
    "# class No value check\n",
    "print(\"class : %d \" % (Y_train[0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "# One - Hot Encoding\n",
    "Y_train = np_utils.to_categorical(Y_train, 10)\n",
    "Y_test = np_utils.to_categorical(Y_test, 10)\n",
    "\n",
    "# print(y_train[0])\n",
    "Y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build Model (Simple)\n",
    "model = Sequential()\n",
    "model.add(Dense(512, input_dim=784, activation='relu'))\n",
    "model.add(Dense(10, activation='softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Model parameters\n",
    "model.compile(loss='categorical_crossentropy',\n",
    "              optimizer='adam',\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "MODEL_DIR = './model/'\n",
    "if not os.path.exists(MODEL_DIR):\n",
    "    os.mkdir(MODEL_DIR)\n",
    "\n",
    "modelpath=\"./model/{epoch:02d}-{val_loss:.4f}.hdf5\"\n",
    "checkpointer = ModelCheckpoint(filepath=modelpath, monitor='val_loss', verbose=1, save_best_only=True)\n",
    "early_stopping_callback = EarlyStopping(monitor='val_loss', patience=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Epoch 00001: val_loss improved from inf to 0.15141, saving model to ./model\\01-0.1514.hdf5\n",
      "\n",
      "Epoch 00002: val_loss improved from 0.15141 to 0.10509, saving model to ./model\\02-0.1051.hdf5\n",
      "\n",
      "Epoch 00003: val_loss improved from 0.10509 to 0.08067, saving model to ./model\\03-0.0807.hdf5\n",
      "\n",
      "Epoch 00004: val_loss improved from 0.08067 to 0.07738, saving model to ./model\\04-0.0774.hdf5\n",
      "\n",
      "Epoch 00005: val_loss improved from 0.07738 to 0.06794, saving model to ./model\\05-0.0679.hdf5\n",
      "\n",
      "Epoch 00006: val_loss improved from 0.06794 to 0.06584, saving model to ./model\\06-0.0658.hdf5\n",
      "\n",
      "Epoch 00007: val_loss improved from 0.06584 to 0.06303, saving model to ./model\\07-0.0630.hdf5\n",
      "\n",
      "Epoch 00008: val_loss did not improve from 0.06303\n",
      "\n",
      "Epoch 00009: val_loss improved from 0.06303 to 0.06131, saving model to ./model\\09-0.0613.hdf5\n",
      "\n",
      "Epoch 00010: val_loss did not improve from 0.06131\n",
      "\n",
      "Epoch 00011: val_loss did not improve from 0.06131\n",
      "\n",
      "Epoch 00012: val_loss did not improve from 0.06131\n",
      "313/313 [==============================] - 1s 2ms/step - loss: 0.0669 - accuracy: 0.9795\n",
      "\n",
      " Test Accuracy: 0.9795\n"
     ]
    }
   ],
   "source": [
    "# Training\n",
    "history = model.fit(X_train, Y_train, validation_data=(X_test, Y_test), epochs=30, batch_size=200, verbose=0, callbacks=[early_stopping_callback,checkpointer])\n",
    "\n",
    "# Accuracy\n",
    "print(\"\\n Test Accuracy: %.4f\" % (model.evaluate(X_test, Y_test)[1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Test Loss\n",
    "y_vloss = history.history['val_loss']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.15141184628009796,\n",
       " 0.10508543998003006,\n",
       " 0.08067195117473602,\n",
       " 0.07737606018781662,\n",
       " 0.06794262677431107,\n",
       " 0.06584016978740692,\n",
       " 0.0630340576171875,\n",
       " 0.06553047150373459,\n",
       " 0.0613052137196064,\n",
       " 0.06419933587312698,\n",
       " 0.06263051927089691,\n",
       " 0.06689702719449997,\n",
       " 0.06303255259990692,\n",
       " 0.06354296207427979,\n",
       " 0.062238823622465134,\n",
       " 0.07480233907699585,\n",
       " 0.08101486414670944,\n",
       " 0.06804490089416504,\n",
       " 0.07043122500181198]"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_vloss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Train Loss\n",
    "y_loss = history.history['loss']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.30191582441329956,\n",
       " 0.12504953145980835,\n",
       " 0.08148521929979324,\n",
       " 0.05840423330664635,\n",
       " 0.04500669613480568,\n",
       " 0.03394630551338196,\n",
       " 0.02545815519988537,\n",
       " 0.019814003258943558,\n",
       " 0.014428107999265194,\n",
       " 0.011991948820650578,\n",
       " 0.009579969570040703,\n",
       " 0.006573781371116638,\n",
       " 0.004951903596520424,\n",
       " 0.004329117015004158,\n",
       " 0.0031558407936245203,\n",
       " 0.0034500397741794586,\n",
       " 0.009509565308690071]"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEGCAYAAAB/+QKOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd3iUVfbA8e9JQhJKAAVEilQpCaA0hSwqRNaCsuIPRXERBWQRRazY1rU3Vl3XVVFEBHRlZUUs7MIuKMKCghSRKr0IAZWiBKKQkOT8/rgzZAiTMDOZyaScz/O8zzvlvTNnUt4zt7z3iqpijDHGFBQT7QCMMcaUTpYgjDHG+GUJwhhjjF+WIIwxxvhlCcIYY4xfcdEOIJxq166tTZo0CansL7/8QtWqVcMbUBhYXMGxuIJjcQWnPMb19ddf71PVOn6fVNVys3Xq1ElDNXfu3JDLRpLFFRyLKzgWV3DKY1zAMi3knGpNTMYYY/yyBGGMMcYvSxDGGGP8Kled1MaY0uHo0aOkp6dz5MiRkMrXqFGDdevWhTmq4ivLcSUmJtKwYUMqVaoU8OtagjDGhF16ejpJSUk0adIEEQm6/KFDh0hKSopAZMVTVuNSVfbv3096ejpNmzYN+HWtickYE3ZHjhyhVq1aISUHE34iQq1atYKu0UU0QYjIpSKyQUQ2i8gDfp7vIyKrRGSFiCwTkfMCLRtOixbB5MmNWLQoku9iTMViyaF0CeX3EbEEISKxwBigF5ACXCciKQUOmwOcrartgSHA+CDKhsWCBdC9O7z1VlN69sSShDHGeESyBnEusFlVt6pqNjAF6ON7gKpmei7UAKgKaKBlw2XBAjh6FFSF7GyYNy8S72KMMWVPJBNEA2Cnz/10z2PHEZH/E5H1wAxcLSLgsuGQlgYxMQBKfDz06BGJdzHGlKT9+/fTvn172rdvz+mnn06DBg2O3c/Ozj5p+Xnz5rFw4cJix3HgwAFee+21Io/Zvn07bdu2LfZ7RUIkRzH5a/A6Yfk6Vf0I+EhELgCeBH4baFkAERkGDAOoW7cu80KoAnTt2oavvz6F559fRVbWwVJVi8jMzAzpM0WaxRWcihZXjRo1OHToUFBlYhYvJu6LL8g57zxyO3cOuryv+Ph4FixYAMAzzzxDtWrVuP322wHIysoiKyuryPKzZs2iWrVqtGvX7rjHc3Nzg4orPT2dV199lYEDBxZ6TGZmJnl5ecX6vIHGdeTIkaB+35FMEOnAGT73GwK7CztYVeeLSHMRqR1MWVUdB4wD6Ny5s/YIoQrQrx8sXAjXXNOROv6nrIqaefPmEcpnijSLKzgVLa5169blD7u8805YsaLoAhkZsGoV5OWREBNDTtu2xJ1ySuHHt28PL70UUCwJCQkkJCSwceNG7r77bjIzM6lduzaTJk2iXr16vPzyy4wdO5a4uDhSUlIYPXo0EydOJDY2lqlTp/LKK69w/vnnA2446fLly7njjjsA1/E7f/58kpKSeP7553n//ffJysri//7v/3j88cd56qmn2LZtG+effz4XXXQRzz///AnxVatWjZiYGJKSkjhy5Ai33HILy5YtIy4ujhdffJG0tDTWrl3L4MGDyc7OJi8vj2nTplG/fn2uueYa0tPTOXr0KI8++ijXXnttkT+LxMREOnToENDPDSKbIJYCLUSkKbAL6A/83vcAETkT2KKqKiIdgXhgP3DgZGXDKTnZ7b/91nVYG2NKWEYG5OW523l5SEYGFJUggqSqjBw5kk8++YQ6derwz3/+k4ceeogJEyYwevRotm3bRkJCAgcOHKBmzZoMHz6catWqMWrUqBNe64UXXmDMmDF069aNzMxMEhMTmT17Nps2bWLJkiWoKldccQXz589n9OjRrFmzhhUnS5AeY8aMAWD16tWsX7+eiy++mI0bNzJ27FjuuOMOBgwYQHZ2Nrm5ucycOZP69eszY8YMDh06RJ735xdGEUsQqpojIrcBs4BYYIKqrhWR4Z7nxwJXATeIyFHgMHCtp9Pab9lIxZriGR+1bp0lCGPCLpBv+osWQc+ekJ0N8fEcGT+eqr/9bdhCyMrKYs2aNVx00UWAa5KpV68eAGeddRYDBgzgyiuv5Morrzzpa3Xr1o27776bAQMG0LdvXxo2bMjs2bOZPXv2sW/nmZmZbNq0iUaNGgUV5xdffMHIkSMBaN26NY0bN2bjxo2kpqby9NNPk56eTt++fWnRogXt2rVj1KhR3H///Vx44YVccsklQb1XICJ6JbWqzgRmFnhsrM/tPwN/DrRspDRsCJUr5/Dtt3ZhuTFRkZoKc+a4YYQ9epAX5k5bVaVNmzYs8jOOfcaMGcyfP5/p06fz5JNPsnZt0d9FH3jgAS6//HJmzpxJ165d+eyzz1BVHnzwQW6++ebjjt2+fXvQcfrz+9//ni5dujBjxgwuueQSxo8fz4UXXsjXX3/NzJkzeeyxx1i8eDGPPPJIUO93MnYlNSACjRv/SimcYsWYiiM1FR580O3DLCEhgb179x5LEEePHmXt2rXk5eWxc+dO0tLSeO655zhw4ACZmZkkJSUV2um7ZcsW2rVrx/3330/nzp1Zv349l1xyCRMmTCAzMxOAXbt2sWfPniJfx58LLriAyZMnA7Bx40Z27NhBq1at2Lp1K82aNeP222/niiuuYNWqVezevZsqVapw/fXXc/vtt7N8+fJi/pROZF+ZPRo3/pXVq6tHOwxjTATExMTwwQcfcPvtt5ORkUFOTg533nknLVu25PrrrycjIwNV5a677qJmzZr87ne/4+qrr+aTTz45rpMa4KWXXmLu3LnExsaSkpJCr169SEhIYN26daR6klu1atV49913ad68Od26daNt27b06tXLbye1r1tvvZXhw4fTrl074uLimDRpEgkJCfzzn//k3XffpVKlSpx++uk88sgjLF26lHvvvZeYmBhiYmIYN25c+H9wha0kVBa34qwoN2zYZgXVAwdCfomIKI8rWEWSxRWcSMX17bffFqv8wYMHwxRJeJX1uPz9XrAV5U6uceNfAayZyRhjPKyJycObIL79Frp2jXIwxphyZ/Xq1SdcMJeQkMDixYujFNHJWYLwOP30wyQkuARhjDHh1q5du4CvhygtrInJIzYWWre2JiZjjPGyBOEjOdlqEMYY42UJwkdKCnz3HfzyS7QjMcaY6LME4SM5GVRhw4ZoR2KMMdFnCcKHd04ma2YypmwrznoQy5YtOzY1eCR8/PHHfHuSk8ygQYP44IMPIhZDoGwUk48zz4S4OOuoNiYaFi06NhUTxZ2KqVatWsdGDD322GMnzMyak5NDXJz/01/nzp3p3Llz8QIowscff0zv3r1JSYnIKsphZQnCR3y8SxJWgzAmfIJcDoKYGGjbtnKRs30HsRzEMYMGDeLUU0/lm2++oWPHjlx77bXceeedHD58mMqVKzNx4kRatWrFvHnzeOGFF/j3v//NY489xo4dO9i6dSs7duxg+PDh3Hfffce9bm5uLjfddBPLli1DRBgyZAh33XUXW7ZsYcSIEezdu5cqVarw5ptv8tNPPzF9+nT+97//8dRTTzFt2jSaN29eZNxz5sxh1KhR5OTkcM455/D666+TkJDAAw88wPTp04mLi6NHjx68/PLLTJ06lccff5zY2Fhq1KjB/Pnzg/shFWAJooCUFDjJZI7GmDArsBwEGRkSzuUgjtm4cSOfffYZsbGxHDx4kPnz5xMXF8dnn33GH//4R6ZNm3ZCmfXr1zN37lwOHTpEy5Ytueuuu6hUqdKx51esWMGuXbtYs2YN4JYZBRg2bBhjx46lRYsWLF68mFtvvZXPP/+cK664gt69e3P11VefNN4jR44waNAg5syZQ8uWLbnhhht4/fXXueGGG/joo49Yv349IsLOnW6F5ieeeIJZs2bRoEGDY3EUhyWIAlJS4JNPICsLEhKiHY0xZV8Iy0EwfvwRfvvbqmGPpV+/fsTGxgKQkZHBjTfeyKZNmxARjh496rfM5ZdffmxVujp16vDjjz/SsGHDY883a9aMrVu3MnLkSC6//HIuvvhiMjMzWbhwIf369Tt23MmWOfVnw4YNNG3alJYtWwJw4403MmbMGG677TYSExMZOnQol19+Od09C9l069aNQYMGcc0119C3b9+g368g66QuIDkZcnNh06ZoR2JMxeFdDuLJJ92+S5fwr44GULVqftJ5+OGHSUtLY82aNfzrX//iyJEjfssk+HxTjI2NJScn57jnTznlFFauXEmPHj0YM2YMQ4cOJS8vj5o1a7JixYpj27oQOje1kPUh4uLiWLJkCVdddRUff/zxsWQwduxYnnrqKXbu3En79u3Zv39/0O/pyxJEAb6ryxljSk4El4PwKyMjgwYNGgAwadKkkF9n37595OXlcdVVV/Hkk0+yfPlyqlevTtOmTZk6dSrgTvQrV64ECGqNiNatW7N9+3Y2b94MwN///ne6d+9OZmYmGRkZXHbZZbz00kusWrUKcGtVdOnShSeeeILatWsfa3oKlSWIAlq1cgsIWUe1MeXbfffdx4MPPki3bt3Izc0N+XV27dpFjx49aN++PYMGDeLZZ58FYPLkybz11lucffbZtGnThk8++QSA/v378/zzz9OhQwe2bNlS5GsnJiYyceJE+vXrR7t27YiJiWH48OEcOnSI3r17c9ZZZ9G9e/dj73nvvffSrl072rZtywUXXMDZZ58d8ucCbD0IL9958Zs1U73mmpBfKqwq2joCxWVxBcfWgwhOWY/L1oMIg5QUa2IyxhhLEH4kJ7vpNgr0RRljTFiNGDHi2BXe3m3ixInRDusYG+bqR0qKG263bRu0aBHtaIwpm1QVEYl2GKXamDFjSuy9tJARUUWxGoQfyclubx3VxoQmMTGR/fv3h3RSMuGnquzfv5/ExMSgylkNwg/fBNGnT3RjMaYsatiwIenp6ezduzek8keOHAn6ZFYSynJciYmJx13gF4iIJggRuRT4GxALjFfV0QWeHwDc77mbCdyiqis9z20HDgG5QI6qRm72rAKqV4eGDa2j2phQVapUiaZNm4Zcft68eXTo0CGMEYVHRYsrYglCRGKBMcBFQDqwVESmq6pvw802oLuq/iwivYBxQBef59NUdV+kYiyKrS5njKnoItkHcS6wWVW3qmo2MAU4rsFGVReq6s+eu18BwdV/IiglBdavz59AzBhjKppIJogGgO913umexwpzE/Afn/sKzBaRr0VkWATiK1Jyslt6tJhXqhtjTJklkRplICL9gEtUdajn/kDgXFUd6efYNOA14DxV3e95rL6q7haR04BPgZGqesLk5p7kMQygbt26naZMmRJSvJmZmVSrVu3Y/VWranDHHR0YPXoVXbr8FNJrhkPBuEoLiys4FldwLK7gFCeutLS0rwvt4y3sEuvibkAqMMvn/oPAg36OOwvYArQs4rUeA0ad7D3DNdWGquq+faqg+pe/hPySYVHRpmgoLosrOBZXcMpjXERpqo2lQAsRaSoi8UB/YLrvASLSCPgQGKiqG30eryoiSd7bwMXAmgjGeoJataBOHeuoNsZUXBEbxaSqOSJyGzALN8x1gqquFZHhnufHAo8AtYDXPFdceoez1gU+8jwWB/xDVf8bqVgLk5JiCcIYU3FF9DoIVZ0JzCzw2Fif20OBoX7KbQWKOU9t8SUnw5QpoOqmADfGmIrEptooQkoKHDgAP/wQ7UiMMabkWYIogq0uZ4ypyCxBFMEm7TPGVGSWIIpQrx7UqGEJwhhTMVmCKIKIrS5njKm4LEGchE3aZ4ypqCxBnERKCuzZA/v3RzsSY4wpWZYgTsLbUW3NTMaYisYSxEl4h7paM5MxpqKxBHESjRpBlSpWgzDGVDyWIE4iJgZat7YahDGm4rEEEQAb6mqMqYgsQQQgOdmtLHfwYLQjMcaYkmMJIgDejur166MbhzHGlCRLEAGwSfuMMRWRJYgANGsG8fHWUW2MqVgsQQQgLg5atrQEYYypWCxBBCg52ZqYjDEViyWIAKWkwNatcPhwtCMxxpiSYQkiQCkpbm3qjRujHYkxxpQMSxABstXljDEVjSWIALVs6abdsARhjKkoLEEEKCEBmje3jmpjTMVhCSIIKSlWgzDGVByWIIKQkgKbNsHRo9GOxBhjIi+iCUJELhWRDSKyWUQe8PP8ABFZ5dkWisjZgZaNhuRkyMmBzZujHYkxxkRexBKEiMQCY4BeQApwnYikFDhsG9BdVc8CngTGBVG2xNnqcsaYiiSSNYhzgc2qulVVs4EpQB/fA1R1oar+7Ln7FdAw0LLR0Lq121tHtTGmIoiL4Gs3AHb63E8HuhRx/E3Af4ItKyLDgGEAdevWZd68eSEFm5mZGVDZunW7MnduBuedVzJZItC4SprFFRyLKzgWV3AiFpeqRmQD+gHjfe4PBF4p5Ng0YB1QK9iyvlunTp00VHPnzg3ouF69VM8+O+S3CVqgcZU0iys4FldwLK7gFCcuYJkWck6NZBNTOnCGz/2GwO6CB4nIWcB4oI+q7g+mbDQkJ8OGDZCbG+1IjDEmsiKZIJYCLUSkqYjEA/2B6b4HiEgj4ENgoKpuDKZstKSkwJEjsH17tCMxxpjIilgfhKrmiMhtwCwgFpigqmtFZLjn+bHAI0At4DURAchR1c6FlY1UrMHwXV2uefPoxmKMMZEUyU5qVHUmMLPAY2N9bg8FhgZatjTwnbSvd+/oxmKMMZFkV1IHqWZNqFfProUwxpR/liBCYKvLGWMqAksQIUhJcQnCjcA1xpjyyRJECFJS4NAh2LUr2pEYY0zkWIIIga0uZ4ypCCxBhMAm7TPGVASWIEJQpw6ceqp1VBtjyjdLECEQsdXljDHlnyWIEHkThI1kMsaUV5YgQpScDD/9BHv3RjsSY4yJDEsQIbKOamNMeWcJIkS+k/YZY0x5ZAkiRA0aQFKS1SCMMeWXJYgQibh+CEsQxpjyyhJEMdikfcaY8swSRDGkpMD338OBA9GOxBhjws8SRDFYR7UxpjyzBFEMNmmfMaY8swRRDE2aQGKiJQhjTPkUUIIQkTtEpLo4b4nIchG5ONLBlXaxsdCqlTUxGWPKp0BrEENU9SBwMVAHGAyMjlhUZYhN2meMKa8CTRDi2V8GTFTVlT6PVWgpKfDdd/DLL9GOxBhjwivQBPG1iMzGJYhZIpIE5EUurLLD21G9fn104zDGmHALNEHcBDwAnKOqvwKVcM1M5cP06TSZOBEWLQq6qE3aZ4wprwJNEKnABlU9ICLXA38CMk5WSEQuFZENIrJZRB7w83xrEVkkIlkiMqrAc9tFZLWIrBCRZQHGGbxZs6BPHxq/8w707Bl0kjjzTIiLs45qY0z5E2iCeB34VUTOBu4DvgPeKaqAiMQCY4BeQApwnYikFDjsJ+B24IVCXiZNVduraucA4wze8uUg4jpUsrNh3rygileqBC1aWA3CGFP+BJogclRVgT7A31T1b0DSScqcC2xW1a2qmg1M8ZQ/RlX3qOpS4GiQcYdPjx6QmIi6gKB796BfwkYyGWPKo7gAjzskIg8CA4HzPbWDSicp0wDY6XM/HegSRGwKzBYRBd5Q1XH+DhKRYcAwgLp16zIvyBoAQPXnn6fe5MnUW7SIdf/6Fz9mZwdVvkqVJmzZ0pjZs+cTHx/eNUgzMzND+kyRZnEFx+IKjsUVnIjFpaon3YDTgbuB8z33GwE3nKRMP2C8z/2BwCuFHPsYMKrAY/U9+9OAlcAFJ4uzU6dOGqq5c+aodumietppqj//HFTZf/xDFVRXrQr57QuPa+7c8L9oGFhcwbG4gmNxBac4cQHLtJBzakBNTKr6AzAZqCEivYEjqlpkHwSuxnCGz/2GwO5A3s/znrs9+z3AR7gmq8iJiYHXXnOLTD/6aFBFvUNdraPaGFOeBDrVxjXAElyt4BpgsYhcfZJiS4EWItJUROKB/sD0AN+vqudaC0SkKu4K7jWBlC2Wjh1h+HB49VVYtSrgYq1auQWErB/CGFOeBNoH8RDuGog9ACJSB/gM+KCwAqqaIyK3AbOAWGCCqq4VkeGe58eKyOnAMqA6kCcid+JGPNUGPhIRb4z/UNX/hvIBg/bUU/D++zBiBMyf7878J1G5MjRrZgnCGFO+BJogYrzJwWM/AdQ+VHUmMLPAY2N9bv+Aa3oq6CBwdoCxhdepp8Lo0fCHP8DkyXD99QEVs9XljDHlTaDDXP8rIrNEZJCIDAJmUODEX64MGQLnngujRkHGSa8HBNxQ1w0bICcnwrEZY0wJCbST+l5gHHAW7pv9OFW9P5KBRVVMDIwZA3v2wGOPBVQkJQWOHoWtWyMbmjHGlJSAFwxS1Wmqereq3qWqH0UyqFKhc2cYNgxeeQVWrz7p4ba6nDGmvCkyQYjIIRE56Gc7JCIHSyrIqHn6aahRA267zV1lXQRLEMaY8qbIBKGqSapa3c+WpKrVSyrIqKlVC5591o1meu+9Ig9NSoKGDa2j2hhTftia1Cdz002uuWnUKDhYdKXJ5mQyxpQnliBOJjbWdVj/8AM8/niRh6akuBpEni2lZIwpByxBBOLcc11N4m9/g7VrCz0sORkOH4YdO0owNmOMiRBLEIF69lmoXr3IDmvv6nKPPRbS4nTGGFOqWIIIVO3a8MwzbkGhf/7T7yGZmW4f4uJ0xhhTqliCCMYf/uAm9LvnHjh06ISnv/nG7VVDWpzOGGNKFUsQwfB2WO/eDU88ccLTPXpAQoK7LeLuG2NMWWUJIlhdu7q5ml566YQxrampMHeuGxWbl+fWqzbGmLLKEkQoRo+GatVg5MgTOqxTU+HTT+H002HwYMjKilKMxhhTTJYgQlGnjpuG4/PPYerUE56uWRPGjYM1a9xhxhhTFlmCCNXNN0OHDnD33fnDl3xcfjkMHOhGx65YEYX4jDGmmCxBhMrbYb1rl1uFzo+XXnKjYwcPdlOBG2NMWWIJojhSU2HQIPjLX2D9+hOePvVUeP11V4MYPbrkwzPGmOKwBFFcf/4zVK3qt8Ma4MoroX9/ePLJgJaVMMaYUsMSRHGddpprYvrsM5g2ze8hr7ziOq4HD7YlSY0xZYcliHAYPhzOPtt1WP/yywlP167tuiu+/hpeeCEK8RljTAgsQYRDXJzLADt3wi23uKFLBSZi6tcPrroKHn3UFhUyxpQNcdEOoNzo1g169YK//x1iYtycG3PmuI5sjzFj3PxMQ4bAF1+4gVDGGFNaWQ0inDp0cPu8PL+z9dWtCy+/DF995YbAGmNMaRbRBCEil4rIBhHZLCIP+Hm+tYgsEpEsERkVTNlSqXfv/Nn6cnMhMfGEQ667Dq64Av70J9i4sYTjM8aYIEQsQYhILDAG6AWkANeJSEqBw34CbgdeCKFs6eOdre+++6B5c7j3Xhg79rhDRNy1EYmJbpE6W57UGFNaRbIGcS6wWVW3qmo2MAXo43uAqu5R1aVAweuMT1q21EpNdddGfPMNXHqp67S++25Xo/CoXx/++lfXDzFmTBRjNcaYIkSyk7oBsNPnfjrQJdxlRWQYMAygbt26zAtxlZ7MzMyQyxbqrrs4MzGRhn/9K/sWL2bdn/5EbuXKADRuDF26tOO++2pSq9ZS6tc/UnJxhYHFFRyLKzgWV3AiFpeqRmQD+gHjfe4PBF4p5NjHgFGhlPXdOnXqpKGaO3duyGVP6tVXVWNiVNu3V92589jDO3aoJiWppqWp5uZGIa5isLiCY3EFx+IKTnHiApZpIefUSDYxpQNn+NxvCOwugbKlz4gRMGMGbNkCXbq4K+aAM85w0zjNneumBzfGmNIkkgliKdBCRJqKSDzQH5heAmVLp0svhYUL3UV1F1wAH38MwNCh8Nvfuv7s776LcozGGOMjYglCVXOA24BZwDrgfVVdKyLDRWQ4gIicLiLpwN3An0QkXUSqF1Y2UrGWmLZtYfFit+/bF/7yFwTlzTfdPH9/+IPf+f6MMSYqInoltarOBGYWeGysz+0fcM1HAZUtF04/3V1Ad+ONMGoUbNhAkzFjeO65SowYARMmuOGvxhgTbXYldTRUrgxTpsBDD8Gbb0KvXgzvf4AePdyI2PT0aAdojDGWIKInJsZNEz5pEsyfT0y3VMY/soOcHLeaqTU1GWOizRJEtN14I3z6KezZQ/NrOvHM0K3MnOnm/DPGmGiyBFEadO/uZvA75RRGvp5Ct5Z7uOMO+P77aAdmjKnILEGUFi1awFdfEfObrkzYeB5HMo9y7bXK5MmNCi4tYYwxJcISRGly6qkwezYtB3Xjppw3WLBAeGt8E3qm5VqSMMb4t2gRjSZPPmGRsnCwBFHaxMfDhAnUO+cMIA8lhiNZ8Pnv34SHH3ajn9ascetNGGMqJlX44Qd47TXo3p2mb70FPXuGPUnYinKlkQgXdsqg8tIjHCERJZZPdnbilmf+yKl5+9wxcXHQsqW76K5NG7dv29ZNM25L1RlTPmRnw+bNsH49bNhw/D4j49hh4j123rzjVrEsLksQpVTqDS2YM+Ey5h79DYdiTuHFmHs4p+EePn5+E+2ylsHata4msWwZTJ2aPy42IQGSk49PHG3awO7dMH8+9OgR1j8gY0wIFi1yJ/MePaBrV9i3L//E75sEtm07bqkAGjSAVq1gwAC3z8uDBx8kLzubmPh493phZAmitEpNJXXes9SdMIFmQ35HH4mhb1/oekNLJk1qSb+nfY795RdYt84lDG/imDcP3n33xNeNi3PrVdx0E9SoUVKfxhjjtWCBm4Dt6FG3gljVqnDoUP7zCQmudaB9e+jfH1q3dsmgVStISjrx9bp0YfuECTQbMiTsX/4sQZRmqansyMqiWWoqXXGTwF51FVxzDTzwgLvOLjYW9wfWubPbfB04AN9+C88/D5984moZOTlwzz1umo+zzoLzzoPzz3f7Bg2i8SmNqRj27HHTNj/3XH4foqqr8XsTQevW0KhRcM3EPueJcLMEUYbUq+emBr/9dhg9GlasgH/8A045pZACNWvCb37jlkCdNcv9UcbHuxrETz+5Je0mTcpf1q5Jk/xkcd557g9XpIQ+nTHl1NdfwyuvwHvvuf/BLl3cipO5ue7/8aWXSm2zryWIMiYhAd54Azp2hOgFlqUAABdbSURBVJEj4ZxzXOWgTZsiCqWmwpw5+W2evn+MR4/CypUuWXzxhUsk3su4a9WCbt3yaxkdO7o/aO+wuoSEUvuHbUxUHT0KH34IL7/spvmvVs1N13zbba6W4NsHUYr/hyxBlFE33+z6n6+6yn0heecdN4N4oVJT/f8hVqqU3zx1552uyrt5c37CWLAApnuW4khMdH/ca9bQNDfXJZJp0+Cyy6ymUVLKyImlwtq71zUjvfaaGxjSvLmrIQwadHyfX2H/j6WMJYgyrFs3V3vt29cliocegieecPMAhkzEXdXdogUMHuwe+/FH+PJLlzDefx9yctywuqws6N3bzU7bpInbGjc+8XbdupZAikPV/Q5mzIBbb3X9SHFx8Ne/utpdjRpuq169mL/8YqjoiWv58vxmpKwsuPhilyh69Yre7yQMLEGUcQ0awP/+51Y1ffpp1y8xeXKYByjVreuyUN++0K8f9OyJZmUhcXHuhBUTA9u3u23JEti///jyiYkuWfgmD28C2bfPjbpKS6uYJxZwzRG7drklBf1tO3a4k46v7Gz3S/cl4ka51KyZnzS8twvua9SA3btpOWeOG/mWkuISUSjbunXw6KMuccXHuzbPiy4qHV8KIpm4jh6Fjz5yzUhffukGiwwdmt+MVA5YgigHEhNh/HjXRXDnnXDuuW5F0+TkCLyZpz9jW1HD6g4dcie27dvz997tm29cNbwgEXdSueACF3hyMpx5pmsCK8tUYc4cmr/xhvvsVauemAB27XLj2X3VresSaPv20KePu334sLua3luDeO459w3hwAF30ZR373t71y6XALyPFXif+uBqJuFy5AhccglUqeJiK2xr2NAtnhXn5xQUah9XXh5kZrq/v0OH3El7xAh3Iq9Uyf2TdOvmkmi1au71g0li3rh+/dV9E3vtNffzbdbM1eYGDy53Q8ctQZQTIu5/oV079yW/Sxd3GcQVV0TgzU42rC4pKf/Kbn9++cWdGJ97znWeeL+JLl4Ms2fnHxcX55KEN2F4t9at3Yk2kgr75pmb60aA7d3rtn37/N/23v/xR8jJ4QyADz7I/1wNG7qTflpafu2qUaP8fWKi/7jOOy/0b8Sq7mefkeGGPr/yijupxsTAkCFw/fXuDynYbdUqGD7cnYjj4mDYMHdC3rXLbV9+6drjC04PExPjEqFv4sjNhYkTaXr0KLz9tquh1q7tTvgHD+af/P3dz8ws/LNnZcHAgcc/FhvrEoXv5k0eBbeffoI333RxjR/vyl90EYwd65qRyunsBZYgypkLLnAXV/ft6754PvooPPJIKWsGrVrVNWncfLPr0/AOv/3Pf1yGW7/eNVt4t2+/dR3lvleUNmqUnzBSUtw+M5NGU6e6k97ZZ7uTQlaW+1Zb1L7gY1u2uA743Fz3g2vb1j23d687URS2mlONGu5kVqeOi69jR/daCxa4MjEx7vqTZ54J/YRSnM5NkfwT3rXXwrhx5GVlEZOQ4BJEqK/bsaO7iKuoxKXqEqY3aRTctmxxV/r//LMLFVzC+dvfXPmEBHfyrl7d7ZOS4LTT3BcI733v5j1m1y7405/yaxAPP+ySUGZmfk3De9t327XrxMc8v3Px/hzvuMPVGso5SxDl0BlnuP+1W26Bxx93LRsjR8LSpaWsD7Gw4bf+LvrzzknjmzjWrXMf9PDhY4c1A7ewd7jk5rpvqp0755/869Q58Xbt2i7JFbRoEfTsmX8ivvLK0vFt0/OzD9sVuCdLXCL5P6/27Qs/bu5cuOwyNDsbiY93baUXXhh6U2O3bsXvg1B1r3HZZW5Ki4QEd7VqBWAJopyqXBkmTnRf7u66C/71L/c/mpDgzsmlKkkEEkx8vKsppKQc/3henuvEfeyx/OYqEbj8cvjd79wHTkw8fu/vMd/90qVuKgRvzWby5NB/YOE+EYdTBK/ADVlaGnz+edF9XMEIx3BSkWNxlcrfYwRZgijHRNxV16tXu2ZTVfdle9IkNz9YaRhkUmwxMW5ElKe56tg39T/+MfR/4t/8pvALC0NRGk/EpVlp/XmV1rgiqDS1TJsIGTLE1Si8CWHcOFfz/uSTEwfPlFneb+pDhoSnipSaCg8+WGG+KRrjj9UgKgDfpv6uXd2ox7/8xTWHJye7qZp+/3v/TehlSgX8hmdMJEW0BiEil4rIBhHZLCIP+HleRORlz/OrRKSjz3PbRWS1iKwQkWWRjLMi8H4hTktz1/Fs2uSa1uPi3PDt5s3hxRePn3XYGFOxRSxBiEgsMAboBaQA14lIgR5GegEtPNsw4PUCz6epantVLTCkxRRXXJyrNaxcCTNnugRxzz1uGP7DD/u/ls0YU7FEsgZxLrBZVbeqajYwBehT4Jg+wDvqfAXUFJF6EYzJFCDirvOZN8+NyOze3a0z0aiRq2ls2xbtCI0x0SJa2EU/xX1hkauBS1V1qOf+QKCLqt7mc8y/gdGq+oXn/hzgflVdJiLbgJ8BBd5Q1XGFvM8wXO2DunXrdpoyZUpI8WZmZlKtWrWQykZSNOLasaMKU6acwaef1iUvT0hL20P//js488xfohpXICyu4FhcwSmPcaWlpX1daCuNqkZkA/oB433uDwReKXDMDOA8n/tzgE6e2/U9+9OAlcAFJ3vPTp06aajmzp0bctlIimZcO3eq3nOParVqbi6MSy9VnTtX9csvVYcO3aILF0YttELZ7zE4FldwymNcwDIt5JwaySamdHBT0Hg0BHYHeoyqevd7gI9wTVamBDVsCC+84K5De/ppN7V4WppbO+itt5rSs6drljLGlE+RTBBLgRYi0lRE4oH+wPQCx0wHbvCMZuoKZKjq9yJSVUSSAESkKnAxsCaCsZoinHKKu+7su+/c5H95eaAqHD4Md9/tJrY0xpQ/EUsQqpoD3AbMAtYB76vqWhEZLiLDPYfNBLYCm4E3gVs9j9cFvhCRlcASYIaq/jdSsZrAVK4MDzzgvehOiY11tYoOHdyUHmPGHJtrzRhTDkT0OghVnamqLVW1uao+7XlsrKqO9dxWVR3heb6dqi7zPL5VVc/2bG28ZU30eS+6u+mmbSxYAD/84GaNVnWjnurVc8NnP/usHF2lbUwFZVNtmKClpsKAATtITYVTT3WJ4Ztv3KqLf/gD/Pe/bqr8Zs3cbLLffRftiI0xobAEYcKmQwdXm9i92y3N26KFSxBNm7oleqdMccstGGPKBksQJuwSE6F/f/j0U9i61S1YtGEDXHcd1K/v1qawjm1jSj9LECaimjRxSzVs2+ZWE73kEnjzzfyO7XvucYt+2XBZY0ofSxCmRMTEuH6J995zTVCvvOJWcnzxRXeNxXnnuWWNV60qfEVPY0zJsgRhSpy3Y3vw4Py1svPy4I033FLSjRu75VJnzIBff41urMZUZJYgTNT06OFW+YyNdddWTJ/uVr7r3Bn+/nfo3Rtq1XL71193V3QbY0qOLRhkosZ3ISPflT1vugmysmD+fPj3v902Y4Z7rl07lzAuv9wtfhQbG63ojSn/LEGYqCpsTfmEBNdncdFF8NJLbhSUN1E89xw8+6yrXfTq5ZLFJZfA+vUweXIjEhJspVBjwsEShCn1RKB1a7eNGgUHDsCsWS5ZzJwJ777r25fRlHfegalT4Xe/y1+H2xgTPOuDMGVOzZpw7bXwzjvw44+wcCFccIF3ag8hOxv69HHTflx2mRtG+9FH7opuGyFlTOCsBmHKtNhY15z0zDPQsydkZeVRqVIMt94KP/3kpv+YPRtyc93xp57qrr/o1MntO3Z0U4LE2FclY05gCcKUC94O7wkTtjNkSLPj+iAOH4bVq93Ms8uXu+3FF+HoUfd89er5ycK77d8PCxYc33luTEVjCcKUG6mpkJW1g9TUZsc9XrkynHuu27yysmDtWpcsvIljzBj3uK+4OHe191VXQdu27rWMqSgsQZgKKSEhv7YwdKh77OhRNxLq8cfhww9df0VODvz5z26LiXEd5e3bH7/VqRPdz2JMpFiCMMajUiV3ncU997jRUdnZEB8Pb7/t+jpWrHDbggXwj3/kl6tf/8Sk0bx5fr/GokU2/NaUTZYgjCmgsAv4+vbNP2b/fli5Mj9prFjhht56O8OrVnXThpx+OvzrX5CT05R334Vp09w1G3aBnykLLEEY40dhF/B51aoFF17oNq8jR+Dbb49PGjNmeDvDhSNH3EV9MTGu/Gmn5W916hR+u2bN/Os5Fi06MXEZEymWIIwJk8TE/H4Nr4ULvcNvlUqVhBEjoFo12LPHbXv3utX49uxxFwD6U6mSSxRVqrj1NfLyXOf50KFuuK43oXj3SUl2gaAJD0sQxkTQb34Dn38OEyZsO2H4bUHZ2bBvX37i8E0ie/bAl1/mr/OdkwNjx/p/nfj4/IRRMHn47tPT4d//bmx9I2XcwoXw1ltNIvJ7tARhTIQVNvy2oPh41+Fdv77/5xctcrURb+f5zJmuM7xgIim437jR7f1Pnd6Ut992CzudeSY0aHD8Vr++29eta/0mpcW+fbB0KSxZ4vq93GJbjZk2zfWdhTNJWIIwpoworPP8jDMCK//LLy5p7N0Lr77qplRXdc1RVarAwYOwbh388EN+Z7tXbKzrcPcmjIIJZO9e1/9y8cVw/vnh/NQV26+/uibIJUvyt61b3XMiriYoAqpuipl58yxBGFNhnazzvChVq7qtSRO3et/UqW5qkoSEGMaPz3/d3FxX49i1y227d+ff3rULNm1yJyJ/fSZPPeVqN6ecAjVquKvUvXvf20U9t2EDTJnSiPh410RXHGWpUz8nxyXZJUvyawirV+cn6zPOcBd73nyz23fqBGvW5E8xEx8fQ48e4Y3JEoQxFVBRU5PExrqJDuvVc4s3FeaXX+D7793062+95fpHRKBLF0hOhowMVyvJyHCTKnrvHzwYyKSJzZgwwXXGJyS4pFOpktv73i64972dkeH6f/Ly3GcaMABatXK1papV/e8LPlap0vEjyIp7PYuqO+EvXOimr69e3SXaJUvcFf3eZsCaNV0SePBBtz/nHFeDK6io32M4RDRBiMilwN+AWGC8qo4u8Lx4nr8M+BUYpKrLAylrjCmeQPtGClO1quu3GDzYTbnu7Rv585+LPoHm5bnk4k0evvv33oOPP85v+urWzX1Tzs52w4X97b23Dx92r+G979tUlpPjLngMVmys+5xxcfDzz6DalLfeck1rlSq5z5Kb67aT3c7L858YK1Vyn3Ho0PwpYc48M/CRaMX9PRYlYglCRGKBMcBFQDqwVESmq+q3Pof1Alp4ti7A60CXAMsaY0qBwvpGChMT44biJiW5/gtfDRvCf/+b3/T17LOhf1sv2Kn/2WduCPKvv7oE9euvx98uar9woWv2AXfWrl0bzjrLJZCYGLf3ve3vMe/tL790Py9V99gjj7gp6UujSNYgzgU2q+pWABGZAvQBfE/yfYB3VFWBr0SkpojUA5oEUNYYU0oUp2+k4OuEq8mksMSVmOimfQ+GN9l4E9frrxcvcX35ZX7i6tkztNcpCaIRWkFFRK4GLlXVoZ77A4EuqnqbzzH/Bkar6hee+3OA+3EJosiyPq8xDBgGULdu3U5TpkwJKd7MzEyqVasWUtlIsriCY3EFx+IK3Nq11VmypDLnnnuYNm0OFvu1VqyoSfv2B4r9WlC8n1daWtrXquq3tymSNQh/LWgFs1FhxwRS1j2oOg4YB9C5c2ftEWI3/rx58wi1bCRZXMGxuIJjcQWuRw9o0yY8cYX7o0Xq5xXJBJEO+I7QbgjsDvCY+ADKGmOMiaBILrS4FGghIk1FJB7oD0wvcMx04AZxugIZqvp9gGWNMcZEUMRqEKqaIyK3AbNwQ1UnqOpaERnueX4sMBM3xHUzbpjr4KLKRipWY4wxJ4rodRCqOhOXBHwfG+tzW4ERgZY1xhhTciLZxGSMMaYMswRhjDHGr4hdBxENIrIX+C7E4rWBfWEMJ1wsruBYXMGxuIJTHuNqrKp1/D1RrhJEcYjIssIuFokmiys4FldwLK7gVLS4rInJGGOMX5YgjDHG+GUJIt+4aAdQCIsrOBZXcCyu4FSouKwPwhhjjF9WgzDGGOOXJQhjjDF+VfgEISKXisgGEdksIg9EOx4AETlDROaKyDoRWSsid0Q7Jl8iEisi33jW8ygVPItNfSAi6z0/t1KxRL2I3OX5Ha4RkfdEJDGKsUwQkT0issbnsVNF5FMR2eTZn1JK4nre87tcJSIfiUjN0hCXz3OjRERFpHZpiUtERnrOZWtF5LlwvFeFThA+S5v2AlKA60QkJbpRAZAD3KOqyUBXYEQpicvrDmBdtIMo4G/Af1W1NXA2pSA+EWkA3A50VtW2uIkn+0cxpEnApQUeewCYo6otgDme+yVtEifG9SnQVlXPAjYCD5Z0UPiPCxE5A7cc8o6SDshjEgXiEpE03KqbZ6lqG+CFcLxRhU4Q+CyLqqrZgHdp06hS1e9Vdbnn9iHcya5B0aVKhog0BC4Hxkc7Fi8RqQ5cALwFoKrZqnogulEdEwdUFpE4oApRXNdEVecDPxV4uA/wtuf228CVJRoU/uNS1dmqmuO5+xVuTZiox+XxV+A+ClnELNIKiesW3OqcWZ5j9oTjvSp6gmgA7PS5n04pORF7iUgToAOwOLqRHPMS7p8jL9qB+GgG7AUmepq+xotI1WgHpaq7cN/kdgDf49Y7mR3dqE5Q17MGC579aVGOx58hwH+iHQSAiFwB7FLVldGOpYCWwPkislhE/ici54TjRSt6ggh4adNoEJFqwDTgTlUt/sK1xY+nN7BHVb+OdiwFxAEdgddVtQPwC9FpKjmOpz2/D9AUqA9UFZHroxtV2SIiD+GaXCeXgliqAA8Bj0Q7Fj/igFNwTdL3Au+LiL/zW1AqeoIIZFnUqBCRSrjkMFlVP4x2PB7dgCtEZDuuOe5CEXk3uiEB7veYrqreWtYHuIQRbb8FtqnqXlU9CnwI/CbKMRX0o4jUA/Dsw9I0EQ4iciPQGxigpeOCrea4ZL/S8z/QEFguIqdHNSonHfhQnSW4Gn6xO9AreoIolUubejL/W8A6VX0x2vF4qeqDqtpQVZvgflafq2rUvxGr6g/AThFp5XmoJ/BtFEPy2gF0FZEqnt9pT0pB53kB04EbPbdvBD6JYizHiMilwP3AFar6a7TjAVDV1ap6mqo28fwPpAMdPX9/0fYxcCGAiLQE4gnDrLMVOkF4OsG8S5uuA94vJUubdgMG4r6hr/Bsl0U7qFJuJDBZRFYB7YFnohwPnhrNB8ByYDXu/y1qUzWIyHvAIqCViKSLyE3AaOAiEdmEG5kzupTE9SqQBHzq+fsfW+SLlFxcUVdIXBOAZp6hr1OAG8NR67KpNowxxvhVoWsQxhhjCmcJwhhjjF+WIIwxxvhlCcIYY4xfliCMMcb4ZQnCmFJARHqUptlxjQFLEMYYYwphCcKYIIjI9SKyxHPx1huetTEyReQvIrJcROaISB3Pse1F5CufNQ1O8Tx+poh8JiIrPWWae16+ms+aFpPDMZeOMcVhCcKYAIlIMnAt0E1V2wO5wACgKrBcVTsC/wMe9RR5B7jfs6bBap/HJwNjVPVs3NxM33se7wDciVubpBnuinpjoiYu2gEYU4b0BDoBSz1f7ivjJrfLA/7pOeZd4EMRqQHUVNX/eR5/G5gqIklAA1X9CEBVjwB4Xm+JqqZ77q8AmgBfRP5jGeOfJQhjAifA26p63OpmIvJwgeOKmr+mqGajLJ/budj/p4kya2IyJnBzgKtF5DQ4tp5zY9z/0dWeY34PfKGqGcDPInK+5/GBwP8863qki8iVntdI8KwzYEypY99QjAmQqn4rIn8CZotIDHAUGIFboKiNiHwNZOD6KcBNnz3WkwC2AoM9jw8E3hCRJzyv0a8EP4YxAbPZXI0pJhHJVNVq0Y7DmHCzJiZjjDF+WQ3CGGOMX1aDMMYY45clCGOMMX5ZgjDGGOOXJQhjjDF+WYIwxhjj1/8DuADcIzPUb0AAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize\n",
    "x_len = numpy.arange(len(y_loss))\n",
    "plt.plot(x_len, y_vloss, marker='.', c=\"red\", label='Test set_loss')\n",
    "plt.plot(x_len, y_loss, marker='.', c=\"blue\", label='Train set_loss')\n",
    "\n",
    "plt.legend(loc='upper right')\n",
    "plt.grid()\n",
    "plt.xlabel('epoch')\n",
    "plt.ylabel('loss')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## MNIST (Deep)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Owner\\miniconda3\\envs\\jeff_heaton_t81_park\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "C:\\Users\\Owner\\miniconda3\\envs\\jeff_heaton_t81_park\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "C:\\Users\\Owner\\miniconda3\\envs\\jeff_heaton_t81_park\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "C:\\Users\\Owner\\miniconda3\\envs\\jeff_heaton_t81_park\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "C:\\Users\\Owner\\miniconda3\\envs\\jeff_heaton_t81_park\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "C:\\Users\\Owner\\miniconda3\\envs\\jeff_heaton_t81_park\\lib\\site-packages\\tensorboard\\compat\\tensorflow_stub\\dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n"
     ]
    }
   ],
   "source": [
    "#-*- coding: utf-8 -*-\n",
    "\n",
    "from keras.datasets import mnist\n",
    "from keras.utils import np_utils\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPooling2D\n",
    "from keras.callbacks import ModelCheckpoint,EarlyStopping\n",
    "import numpy\n",
    "import sys\n",
    "import os\n",
    "import tensorflow as tf\n",
    "\n",
    "# set seed \n",
    "seed = 0\n",
    "numpy.random.seed(seed)\n",
    "tf.random.set_seed(seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "(X_train, Y_train), (X_test, Y_test) = mnist.load_data()\n",
    "X_train = X_train.reshape(X_train.shape[0], 28, 28, 1).astype('float32') / 255\n",
    "X_test = X_test.reshape(X_test.shape[0], 28, 28, 1).astype('float32') / 255\n",
    "Y_train = np_utils.to_categorical(Y_train)\n",
    "Y_test = np_utils.to_categorical(Y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 28, 28, 1)\n",
      "(10000, 28, 28, 1)\n",
      "(60000, 10)\n",
      "(10000, 10)\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape)\n",
    "print(X_test.shape)\n",
    "print(Y_train.shape)\n",
    "print(Y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Epoch 00001: val_loss improved from inf to 0.05721, saving model to ./model\\01-0.0572.hdf5\n",
      "\n",
      "Epoch 00002: val_loss improved from 0.05721 to 0.04122, saving model to ./model\\02-0.0412.hdf5\n",
      "\n",
      "Epoch 00003: val_loss improved from 0.04122 to 0.03531, saving model to ./model\\03-0.0353.hdf5\n",
      "\n",
      "Epoch 00004: val_loss improved from 0.03531 to 0.03311, saving model to ./model\\04-0.0331.hdf5\n",
      "\n",
      "Epoch 00005: val_loss improved from 0.03311 to 0.02918, saving model to ./model\\05-0.0292.hdf5\n",
      "\n",
      "Epoch 00006: val_loss improved from 0.02918 to 0.02905, saving model to ./model\\06-0.0291.hdf5\n",
      "\n",
      "Epoch 00007: val_loss improved from 0.02905 to 0.02770, saving model to ./model\\07-0.0277.hdf5\n",
      "\n",
      "Epoch 00008: val_loss did not improve from 0.02770\n",
      "\n",
      "Epoch 00009: val_loss did not improve from 0.02770\n",
      "\n",
      "Epoch 00010: val_loss did not improve from 0.02770\n",
      "\n",
      "Epoch 00011: val_loss did not improve from 0.02770\n",
      "\n",
      "Epoch 00012: val_loss did not improve from 0.02770\n",
      "\n",
      "Epoch 00013: val_loss did not improve from 0.02770\n",
      "\n",
      "Epoch 00014: val_loss did not improve from 0.02770\n",
      "\n",
      "Epoch 00015: val_loss did not improve from 0.02770\n",
      "\n",
      "Epoch 00016: val_loss did not improve from 0.02770\n",
      "\n",
      "Epoch 00017: val_loss did not improve from 0.02770\n",
      "313/313 [==============================] - 3s 9ms/step - loss: 0.0294 - accuracy: 0.9925\n",
      "\n",
      " Test Accuracy: 0.9925\n"
     ]
    }
   ],
   "source": [
    "# Build Model (Deep)\n",
    "model = Sequential()\n",
    "model.add(Conv2D(32, kernel_size=(3, 3), input_shape=(28, 28, 1), activation='relu'))\n",
    "model.add(Conv2D(64, (3, 3), activation='relu'))\n",
    "model.add(MaxPooling2D(pool_size=2))\n",
    "model.add(Dropout(0.25))\n",
    "model.add(Flatten())\n",
    "model.add(Dense(128,  activation='relu'))\n",
    "model.add(Dropout(0.5))\n",
    "model.add(Dense(10, activation='softmax'))\n",
    "\n",
    "model.compile(loss='categorical_crossentropy',\n",
    "              optimizer='adam',\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "MODEL_DIR = './model/'\n",
    "if not os.path.exists(MODEL_DIR):\n",
    "    os.mkdir(MODEL_DIR)\n",
    "\n",
    "modelpath=\"./model/{epoch:02d}-{val_loss:.4f}.hdf5\"\n",
    "checkpointer = ModelCheckpoint(filepath=modelpath, monitor='val_loss', verbose=1, save_best_only=True)\n",
    "early_stopping_callback = EarlyStopping(monitor='val_loss', patience=10)\n",
    "\n",
    "# Training\n",
    "history = model.fit(X_train, Y_train, validation_data=(X_test, Y_test), epochs=30, batch_size=200, verbose=0, callbacks=[early_stopping_callback,checkpointer])\n",
    "\n",
    "# Accuracy\n",
    "print(\"\\n Test Accuracy: %.4f\" % (model.evaluate(X_test, Y_test)[1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "import matplotlib.pyplot as plt\n",
    "# Val_loss\n",
    "y_vloss = history.history['val_loss']\n",
    "\n",
    "# Train Loss\n",
    "y_loss = history.history['loss']\n",
    "\n",
    "x_len = numpy.arange(len(y_loss))\n",
    "plt.plot(x_len, y_vloss, marker='.', c=\"red\", label='Testset_loss')\n",
    "plt.plot(x_len, y_loss, marker='.', c=\"blue\", label='Trainset_loss')\n",
    "\n",
    "plt.legend(loc='upper right')\n",
    "plt.grid()\n",
    "plt.xlabel('epoch')\n",
    "plt.ylabel('loss')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.6 (tensorflow)",
   "language": "python",
   "name": "jeff_heaton_t81_park"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}