Multilayer_perceptron.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Imports\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.model_selection import LeaveOneOut\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Functions\n",
    "\n",
    "'''\n",
    "Display or save an image containing the loss function, the accuracy and the heatmap of the model\n",
    "\n",
    "    Parameters:\n",
    "        clf (MLPClassifier): The trained model\n",
    "        data (array): Data of the file\n",
    "        label (array): Labels of the data\n",
    "        title (string): Title for the image containing the different parameters\n",
    "'''\n",
    "def display(clf, data, label, title):\n",
    "    plt.figure(figsize=(10,10))\n",
    "    plt.suptitle(title)\n",
    "    \n",
    "    plt.subplot(1,2,1)\n",
    "    plt.title(\"Training loss\")\n",
    "    plt.ylabel(\"Value of loss\")\n",
    "    plt.xlabel(\"Epochs\")\n",
    "    y = clf.loss_curve_\n",
    "    x = list(range(1,len(y)+1))\n",
    "    plt.plot(x,y)\n",
    "    \n",
    "    plt.subplot(1,2,2)\n",
    "    plt.title(\"HeatMap\")\n",
    "    plt.ylabel(\"Predicted labels\")\n",
    "    plt.xlabel(\"Real labels\")\n",
    "    y_pred = clf.predict(data)\n",
    "    cm = confusion_matrix(label_target, y_pred)\n",
    "    ax = sns.heatmap(cm, center=True)\n",
    "    ax.set(xlabel=\"Training labels\", ylabel=\"Predicted labels\")\n",
    "    \n",
    "\n",
    "    plt.show()\n",
    "    #plt.savefig(f'./results/{title[:-33]}.png')\n",
    "    #plt.clf()\n",
    "\n",
    "'''\n",
    "Train the model depending of the input parameters\n",
    "\n",
    "    Parameters:\n",
    "        layer (array): Tuples for the number of nodes per layer and the number of intermediate layers\n",
    "        tols (array): Value of tolerance before training stop if not achieve\n",
    "        epochs (array): Number of epochs \n",
    "\n",
    "'''\n",
    "def train_model(layer, tols, epochs) :\n",
    "    loo = LeaveOneOut()\n",
    "    results = np.zeros(len(data_scaled))\n",
    "    for train, test in loo.split(data_scaled):\n",
    "        clf = MLPClassifier(hidden_layer_sizes=layer, random_state=1, max_iter=epochs, tol = tols).fit(data_scaled[train], label_target[train])\n",
    "        results[test] = clf.predict(data_scaled[test])\n",
    "        \n",
    "    accuracy = (accuracy_score(label_target, results)*100)\n",
    "    title = f\"Results for {layer[1]} layers of {layer[0]} nodes, {epochs} epochs and a tol of {tols} \\n Accuracy = {accuracy} %\"\n",
    "    display(clf, data_scaled, label_target, title)\n",
    "\n",
    "'''\n",
    "Train the model by using different parameters\n",
    "\n",
    "'''\n",
    "def train_parameters() :\n",
    "    layer = [(100,1),(150,1),(100,10),(150,10)]\n",
    "    tols = [10e-4, 10e-5, 10e-3, 10e-2]\n",
    "    epochs = [20,50,100]\n",
    "    \n",
    "    for i in layer :\n",
    "        train_model(i, tols[0], epochs[0])\n",
    "    for j in tols :\n",
    "        train_model(layer[0], j, epochs[0])\n",
    "    for k in epochs :\n",
    "        train_model(layer[0], tols[0], k)\n",
    "\n",
    "'''\n",
    "Search the best parameters for this estimator and the data\n",
    "\n",
    "    Returns:\n",
    "        clf.best_params_ (dict): Dictionnary with all the parameters giving the best accuracy for the model\n",
    "'''\n",
    "def search_best_parameter():\n",
    "    \n",
    "    layer = [(25,1),(50,1),(25,10),(50,10)]\n",
    "    tols = [10e-4, 10e-5, 10e-3, 10e-2]\n",
    "    epochs = [20,50,100]\n",
    "    \n",
    "    loo = LeaveOneOut()\n",
    "    results = np.zeros(len(data_scaled))\n",
    "    for train, test in loo.split(data_scaled):\n",
    "        clf = GridSearchCV(\n",
    "            MLPClassifier(),\n",
    "            param_grid= {\n",
    "                'hidden_layer_sizes': layer,\n",
    "                'tol' : tols,\n",
    "                'max_iter' : epochs\n",
    "                },\n",
    "            refit='True',\n",
    "            cv=2,\n",
    "            n_jobs=-1,\n",
    "        )\n",
    "        clf.fit(data_scaled[train], label_target[train])\n",
    "        results[test] = clf.predict(data_scaled[test])\n",
    "    print (\"Accuracy=%f%%\" % (accuracy_score(label_target, results)*100))\n",
    "    return (clf.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "### Main, data_reading\n",
    "\n",
    "Data_Pam = pd.read_csv(\"./data/Data_Pam50.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Data_manipulation\n",
    "\n",
    "sc_X = StandardScaler()\n",
    "\n",
    "label_raw = Data_Pam.iloc[:,1]\n",
    "data_raw = Data_Pam.iloc[:,2:]\n",
    "data_raw_transposed = data_raw.T\n",
    "\n",
    "values_tot = []\n",
    "for j in data_raw_transposed :\n",
    "    values = []\n",
    "    for k in data_raw_transposed[j] :\n",
    "        values.append(k)\n",
    "    values_tot.append(values)\n",
    "data = np.array(values_tot)\n",
    "data_scaled=sc_X.fit_transform(data)\n",
    "\n",
    "target = {'Normal': 0, 'Basal': 1, 'Her2': 2, 'LumA': 3, 'LumB': 4}\n",
    "label_target = np.zeros(len(label_raw))\n",
    "for i in range (len(label_raw)) :\n",
    "    label_target[i] = target[label_raw[i]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (20) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/usr/lib/python3/dist-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (100) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_parameters()"
   ]
  },
  {
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}