4f_ex_1_solution.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Exercise Solution: Neural Networks (Stroke Thromobolysis Dataset) \n",
    "\n",
    "The data loaded in this exercise is for seven acute stroke units, and whether a patient receives clost-busting treatment for stroke.  There are lots of features, and a description of the features can be found in the file stroke_data_feature_descriptions.csv.\n",
    "\n",
    "Train a Neural Network model to try to predict whether or not a stroke patient receives clot-busting treatment.  Use the prompts below to write each section of code.\n",
    "\n",
    "How accurate can you get your model on the test set?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "# sklearn for pre-processing\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "# TensorFlow sequential model\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import backend as K\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import BatchNormalization\n",
    "from tensorflow.keras.layers import Dense\n",
    "from tensorflow.keras.layers import Dropout\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "\n",
    "# Download data\n",
    "# (not required if running locally and have previously downloaded data)\n",
    "\n",
    "download_required = True\n",
    "\n",
    "if download_required:\n",
    "\n",
    "    # Download processed data:\n",
    "    address = 'https://raw.githubusercontent.com/MichaelAllen1966/' + \\\n",
    "                '2004_titanic/master/jupyter_notebooks/data/hsma_stroke.csv'\n",
    "    data = pd.read_csv(address)\n",
    "\n",
    "    # Create a data subfolder if one does not already exist\n",
    "    import os\n",
    "    data_directory ='./data/'\n",
    "    if not os.path.exists(data_directory):\n",
    "        os.makedirs(data_directory)\n",
    "\n",
    "    # Save data to data subfolder\n",
    "    data.to_csv(data_directory + 'hsma_stroke.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Look at an overview of the data using the `describe()` method of Pandas."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Clotbuster given</th>\n",
       "      <th>Hosp_1</th>\n",
       "      <th>Hosp_2</th>\n",
       "      <th>Hosp_3</th>\n",
       "      <th>Hosp_4</th>\n",
       "      <th>Hosp_5</th>\n",
       "      <th>Hosp_6</th>\n",
       "      <th>Hosp_7</th>\n",
       "      <th>Male</th>\n",
       "      <th>Age</th>\n",
       "      <th>...</th>\n",
       "      <th>S2NihssArrivalFacialPalsy</th>\n",
       "      <th>S2NihssArrivalMotorArmLeft</th>\n",
       "      <th>S2NihssArrivalMotorArmRight</th>\n",
       "      <th>S2NihssArrivalMotorLegLeft</th>\n",
       "      <th>S2NihssArrivalMotorLegRight</th>\n",
       "      <th>S2NihssArrivalLimbAtaxia</th>\n",
       "      <th>S2NihssArrivalSensory</th>\n",
       "      <th>S2NihssArrivalBestLanguage</th>\n",
       "      <th>S2NihssArrivalDysarthria</th>\n",
       "      <th>S2NihssArrivalExtinctionInattention</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "      <td>1862.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.403330</td>\n",
       "      <td>0.159506</td>\n",
       "      <td>0.142320</td>\n",
       "      <td>0.154672</td>\n",
       "      <td>0.165414</td>\n",
       "      <td>0.055854</td>\n",
       "      <td>0.113319</td>\n",
       "      <td>0.208915</td>\n",
       "      <td>0.515575</td>\n",
       "      <td>74.553706</td>\n",
       "      <td>...</td>\n",
       "      <td>1.114930</td>\n",
       "      <td>1.002148</td>\n",
       "      <td>0.963480</td>\n",
       "      <td>0.963480</td>\n",
       "      <td>0.910849</td>\n",
       "      <td>0.216971</td>\n",
       "      <td>0.610097</td>\n",
       "      <td>0.944146</td>\n",
       "      <td>0.739527</td>\n",
       "      <td>0.566595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.490698</td>\n",
       "      <td>0.366246</td>\n",
       "      <td>0.349472</td>\n",
       "      <td>0.361689</td>\n",
       "      <td>0.371653</td>\n",
       "      <td>0.229701</td>\n",
       "      <td>0.317068</td>\n",
       "      <td>0.406643</td>\n",
       "      <td>0.499892</td>\n",
       "      <td>12.280576</td>\n",
       "      <td>...</td>\n",
       "      <td>0.930527</td>\n",
       "      <td>1.479211</td>\n",
       "      <td>1.441594</td>\n",
       "      <td>1.406501</td>\n",
       "      <td>1.380606</td>\n",
       "      <td>0.522643</td>\n",
       "      <td>0.771932</td>\n",
       "      <td>1.121379</td>\n",
       "      <td>0.731083</td>\n",
       "      <td>0.794000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>67.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>76.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>83.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 51 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Clotbuster given       Hosp_1       Hosp_2       Hosp_3       Hosp_4  \\\n",
       "count       1862.000000  1862.000000  1862.000000  1862.000000  1862.000000   \n",
       "mean           0.403330     0.159506     0.142320     0.154672     0.165414   \n",
       "std            0.490698     0.366246     0.349472     0.361689     0.371653   \n",
       "min            0.000000     0.000000     0.000000     0.000000     0.000000   \n",
       "25%            0.000000     0.000000     0.000000     0.000000     0.000000   \n",
       "50%            0.000000     0.000000     0.000000     0.000000     0.000000   \n",
       "75%            1.000000     0.000000     0.000000     0.000000     0.000000   \n",
       "max            1.000000     1.000000     1.000000     1.000000     1.000000   \n",
       "\n",
       "            Hosp_5       Hosp_6       Hosp_7         Male          Age  ...  \\\n",
       "count  1862.000000  1862.000000  1862.000000  1862.000000  1862.000000  ...   \n",
       "mean      0.055854     0.113319     0.208915     0.515575    74.553706  ...   \n",
       "std       0.229701     0.317068     0.406643     0.499892    12.280576  ...   \n",
       "min       0.000000     0.000000     0.000000     0.000000    40.000000  ...   \n",
       "25%       0.000000     0.000000     0.000000     0.000000    67.000000  ...   \n",
       "50%       0.000000     0.000000     0.000000     1.000000    76.000000  ...   \n",
       "75%       0.000000     0.000000     0.000000     1.000000    83.000000  ...   \n",
       "max       1.000000     1.000000     1.000000     1.000000   100.000000  ...   \n",
       "\n",
       "       S2NihssArrivalFacialPalsy  S2NihssArrivalMotorArmLeft  \\\n",
       "count                1862.000000                 1862.000000   \n",
       "mean                    1.114930                    1.002148   \n",
       "std                     0.930527                    1.479211   \n",
       "min                     0.000000                    0.000000   \n",
       "25%                     0.000000                    0.000000   \n",
       "50%                     1.000000                    0.000000   \n",
       "75%                     2.000000                    2.000000   \n",
       "max                     3.000000                    4.000000   \n",
       "\n",
       "       S2NihssArrivalMotorArmRight  S2NihssArrivalMotorLegLeft  \\\n",
       "count                  1862.000000                 1862.000000   \n",
       "mean                      0.963480                    0.963480   \n",
       "std                       1.441594                    1.406501   \n",
       "min                       0.000000                    0.000000   \n",
       "25%                       0.000000                    0.000000   \n",
       "50%                       0.000000                    0.000000   \n",
       "75%                       2.000000                    2.000000   \n",
       "max                       4.000000                    4.000000   \n",
       "\n",
       "       S2NihssArrivalMotorLegRight  S2NihssArrivalLimbAtaxia  \\\n",
       "count                  1862.000000               1862.000000   \n",
       "mean                      0.910849                  0.216971   \n",
       "std                       1.380606                  0.522643   \n",
       "min                       0.000000                  0.000000   \n",
       "25%                       0.000000                  0.000000   \n",
       "50%                       0.000000                  0.000000   \n",
       "75%                       2.000000                  0.000000   \n",
       "max                       4.000000                  2.000000   \n",
       "\n",
       "       S2NihssArrivalSensory  S2NihssArrivalBestLanguage  \\\n",
       "count            1862.000000                 1862.000000   \n",
       "mean                0.610097                    0.944146   \n",
       "std                 0.771932                    1.121379   \n",
       "min                 0.000000                    0.000000   \n",
       "25%                 0.000000                    0.000000   \n",
       "50%                 0.000000                    0.000000   \n",
       "75%                 1.000000                    2.000000   \n",
       "max                 2.000000                    3.000000   \n",
       "\n",
       "       S2NihssArrivalDysarthria  S2NihssArrivalExtinctionInattention  \n",
       "count               1862.000000                          1862.000000  \n",
       "mean                   0.739527                             0.566595  \n",
       "std                    0.731083                             0.794000  \n",
       "min                    0.000000                             0.000000  \n",
       "25%                    0.000000                             0.000000  \n",
       "50%                    1.000000                             0.000000  \n",
       "75%                    1.000000                             1.000000  \n",
       "max                    2.000000                             2.000000  \n",
       "\n",
       "[8 rows x 51 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Convert all of the data in the dataframe to type `float`, pull out the X (feature) and y (label) data, and put X and y into Numpy arrays."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data.astype(float)\n",
    "\n",
    "X = data.drop('Clotbuster given', axis=1)\n",
    "y = data['Clotbuster given']\n",
    "\n",
    "X_np = X.values\n",
    "y_np = y.values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Define a function that will MinMax Normalise training and test feature data passed into it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def scale_data(X_train, X_test):\n",
    "    \"\"\"Scale data 0-1 based on min and max in training set\"\"\"\n",
    "\n",
    "    # Initialise a new scaling object for normalising input data\n",
    "    sc = MinMaxScaler()\n",
    "\n",
    "    # Apply the scaler to the training and test sets\n",
    "    train_sc = sc.fit_transform(X_train)\n",
    "    test_sc = sc.fit_transform(X_test)\n",
    "\n",
    "    return train_sc, test_sc"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Define a function that will build a sequential neural network, given a number of features, a number of hidden layers (with a default of 5), a number of neurons per hidden layer (with a default of 64), a dropout rate (with a default of 0), and a learning rate (with a default of 0.003).  The function should also create a single neuron output layer with a Sigmoid activation function, use an Adam optimiser, and a Binary Crossentropy loss function, with accuracy as the performance metric."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_net(number_features,\n",
    "             hidden_layers=5,\n",
    "             hidden_layer_neurones=64,\n",
    "             dropout=0.0,\n",
    "             learning_rate=0.003):\n",
    "\n",
    "    \"\"\"Make TensorFlow neural net\"\"\"\n",
    "\n",
    "    # Clear Tensorflow\n",
    "    K.clear_session()\n",
    "\n",
    "    # Set up neural net\n",
    "    net = Sequential()\n",
    "\n",
    "    # Add hidden hidden_layers using a loop\n",
    "    for i in range(hidden_layers):\n",
    "        # Add fully connected layer with ReLu activation\n",
    "        net.add(Dense(\n",
    "            hidden_layer_neurones,\n",
    "            input_dim=number_features,\n",
    "            activation='relu'))\n",
    "        # Add droput layer\n",
    "        net.add(Dropout(dropout))\n",
    "\n",
    "    # Add final sigmoid activation output\n",
    "    net.add(Dense(1, activation='sigmoid'))\n",
    "\n",
    "    # Compiling model\n",
    "    opt = Adam(learning_rate=learning_rate)\n",
    "\n",
    "    net.compile(loss='binary_crossentropy',\n",
    "                optimizer=opt,\n",
    "                metrics=['accuracy'])\n",
    "\n",
    "    return net"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Split your data into training and test sets.  Decide on an appropriate test data size.  Then scale the feature data using MinMax Normalisation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split data\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    X_np, y_np, test_size = 0.25, random_state=42)\n",
    "\n",
    "# Scale X data\n",
    "X_train_sc, X_test_sc = scale_data(X_train, X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Write a function to calculate accuracy of the model on both training and test sets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def calculate_accuracy(model, X_train_sc, X_test_sc, y_train, y_test):\n",
    "    \"\"\"Calculate and print accuracy of training and test data fits\"\"\"\n",
    "\n",
    "    ### Get accuracy of fit to training data\n",
    "    probability = model.predict(X_train_sc)\n",
    "    y_pred_train = probability >= 0.5\n",
    "    y_pred_train = y_pred_train.flatten()\n",
    "    accuracy_train = np.mean(y_pred_train == y_train)\n",
    "\n",
    "    ### Get accuracy of fit to test data\n",
    "    probability = model.predict(X_test_sc)\n",
    "    y_pred_test = probability >= 0.5\n",
    "    y_pred_test = y_pred_test.flatten()\n",
    "    accuracy_test = np.mean(y_pred_test == y_test)\n",
    "\n",
    "    # Show acuracy\n",
    "    print (f'Training accuracy {accuracy_train:0.3f}')\n",
    "    print (f'Test accuracy {accuracy_test:0.3f}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Write a function to plot training and test set accuracy over time during model fitting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_training(history_dict):\n",
    "    acc_values = history_dict['accuracy']\n",
    "    val_acc_values = history_dict['val_accuracy']\n",
    "    epochs = range(1, len(acc_values) + 1)\n",
    "\n",
    "    fig, ax = plt.subplots()\n",
    "\n",
    "    ax.set_xlabel('Time')\n",
    "    ax.set_ylabel('Accuracy')\n",
    "\n",
    "    ax.plot(epochs, acc_values, color='blue', label='Training acc')\n",
    "    ax.plot(epochs, val_acc_values, color='red', label='Test accuracy')\n",
    "    ax.set_title('Training and validation accuracy')\n",
    "\n",
    "    ax.legend()\n",
    "\n",
    "    fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create a neural network with a number of hidden layers, neurons, dropout rate and learning rate of your choosing.  Run the model for a number of epochs and with a batch size of your choosing (be careful about using large batch sizes unless you've got a CUDA-enabled GPU and TensorFlow is set up to use it)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From c:\\Users\\dan\\anaconda3\\envs\\tf_hsma\\Lib\\site-packages\\keras\\src\\backend\\common\\global_state.py:73: The name tf.reset_default_graph is deprecated. Please use tf.compat.v1.reset_default_graph instead.\n",
      "\n",
      "Epoch 1/100\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\dan\\anaconda3\\envs\\tf_hsma\\Lib\\site-packages\\keras\\src\\layers\\core\\dense.py:88: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
      "  super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.6053 - loss: 0.6310 - val_accuracy: 0.7232 - val_loss: 0.5111\n",
      "Epoch 2/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7871 - loss: 0.4625 - val_accuracy: 0.7682 - val_loss: 0.4507\n",
      "Epoch 3/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 970us/step - accuracy: 0.8193 - loss: 0.4136 - val_accuracy: 0.8176 - val_loss: 0.4210\n",
      "Epoch 4/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 960us/step - accuracy: 0.8137 - loss: 0.3941 - val_accuracy: 0.8112 - val_loss: 0.3919\n",
      "Epoch 5/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8437 - loss: 0.3576 - val_accuracy: 0.8133 - val_loss: 0.4186\n",
      "Epoch 6/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8605 - loss: 0.3330 - val_accuracy: 0.7983 - val_loss: 0.4143\n",
      "Epoch 7/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 916us/step - accuracy: 0.8259 - loss: 0.3623 - val_accuracy: 0.8326 - val_loss: 0.3939\n",
      "Epoch 8/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8528 - loss: 0.3238 - val_accuracy: 0.8326 - val_loss: 0.4427\n",
      "Epoch 9/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 997us/step - accuracy: 0.8667 - loss: 0.2907 - val_accuracy: 0.7983 - val_loss: 0.4531\n",
      "Epoch 10/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8759 - loss: 0.2771 - val_accuracy: 0.8262 - val_loss: 0.5343\n",
      "Epoch 11/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 966us/step - accuracy: 0.8924 - loss: 0.2407 - val_accuracy: 0.8047 - val_loss: 0.5151\n",
      "Epoch 12/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 961us/step - accuracy: 0.8852 - loss: 0.2259 - val_accuracy: 0.8026 - val_loss: 0.5504\n",
      "Epoch 13/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 881us/step - accuracy: 0.9026 - loss: 0.2148 - val_accuracy: 0.8133 - val_loss: 0.6227\n",
      "Epoch 14/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 963us/step - accuracy: 0.9062 - loss: 0.1986 - val_accuracy: 0.8004 - val_loss: 0.8707\n",
      "Epoch 15/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 940us/step - accuracy: 0.9229 - loss: 0.1931 - val_accuracy: 0.8112 - val_loss: 0.7989\n",
      "Epoch 16/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 888us/step - accuracy: 0.9072 - loss: 0.1717 - val_accuracy: 0.7833 - val_loss: 0.6860\n",
      "Epoch 17/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 930us/step - accuracy: 0.9183 - loss: 0.1623 - val_accuracy: 0.7983 - val_loss: 0.8908\n",
      "Epoch 18/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 992us/step - accuracy: 0.9408 - loss: 0.1574 - val_accuracy: 0.7747 - val_loss: 0.6969\n",
      "Epoch 19/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 968us/step - accuracy: 0.9451 - loss: 0.1396 - val_accuracy: 0.7854 - val_loss: 0.8706\n",
      "Epoch 20/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 924us/step - accuracy: 0.9475 - loss: 0.1148 - val_accuracy: 0.7554 - val_loss: 1.1075\n",
      "Epoch 21/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9511 - loss: 0.1186 - val_accuracy: 0.7961 - val_loss: 0.6700\n",
      "Epoch 22/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9314 - loss: 0.1808 - val_accuracy: 0.7918 - val_loss: 0.7345\n",
      "Epoch 23/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9701 - loss: 0.0757 - val_accuracy: 0.7747 - val_loss: 1.1924\n",
      "Epoch 24/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9593 - loss: 0.0831 - val_accuracy: 0.7854 - val_loss: 1.0050\n",
      "Epoch 25/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9634 - loss: 0.1004 - val_accuracy: 0.7554 - val_loss: 1.0066\n",
      "Epoch 26/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9648 - loss: 0.0856 - val_accuracy: 0.7597 - val_loss: 1.0974\n",
      "Epoch 27/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 957us/step - accuracy: 0.9661 - loss: 0.0847 - val_accuracy: 0.7575 - val_loss: 1.1660\n",
      "Epoch 28/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9689 - loss: 0.0635 - val_accuracy: 0.7897 - val_loss: 1.2466\n",
      "Epoch 29/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 997us/step - accuracy: 0.9821 - loss: 0.0489 - val_accuracy: 0.7704 - val_loss: 1.1211\n",
      "Epoch 30/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9667 - loss: 0.0809 - val_accuracy: 0.8219 - val_loss: 1.1547\n",
      "Epoch 31/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 986us/step - accuracy: 0.9958 - loss: 0.0216 - val_accuracy: 0.7811 - val_loss: 1.6401\n",
      "Epoch 32/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 879us/step - accuracy: 0.9728 - loss: 0.0752 - val_accuracy: 0.7833 - val_loss: 1.2428\n",
      "Epoch 33/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9781 - loss: 0.0615 - val_accuracy: 0.7983 - val_loss: 1.0353\n",
      "Epoch 34/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9903 - loss: 0.0351 - val_accuracy: 0.7940 - val_loss: 1.3027\n",
      "Epoch 35/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9939 - loss: 0.0128 - val_accuracy: 0.7961 - val_loss: 1.6230\n",
      "Epoch 36/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9948 - loss: 0.0153 - val_accuracy: 0.7940 - val_loss: 1.5783\n",
      "Epoch 37/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9911 - loss: 0.0221 - val_accuracy: 0.7854 - val_loss: 1.6514\n",
      "Epoch 38/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 958us/step - accuracy: 0.9853 - loss: 0.0537 - val_accuracy: 0.7961 - val_loss: 1.1731\n",
      "Epoch 39/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9764 - loss: 0.0606 - val_accuracy: 0.7768 - val_loss: 1.3468\n",
      "Epoch 40/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9961 - loss: 0.0203 - val_accuracy: 0.7747 - val_loss: 1.4595\n",
      "Epoch 41/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9928 - loss: 0.0239 - val_accuracy: 0.7897 - val_loss: 1.5007\n",
      "Epoch 42/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9935 - loss: 0.0199 - val_accuracy: 0.7554 - val_loss: 1.5526\n",
      "Epoch 43/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9809 - loss: 0.0577 - val_accuracy: 0.7532 - val_loss: 1.4397\n",
      "Epoch 44/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9807 - loss: 0.0596 - val_accuracy: 0.7790 - val_loss: 1.0982\n",
      "Epoch 45/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9876 - loss: 0.0351 - val_accuracy: 0.7790 - val_loss: 1.3248\n",
      "Epoch 46/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 938us/step - accuracy: 0.9907 - loss: 0.0292 - val_accuracy: 0.7554 - val_loss: 1.5941\n",
      "Epoch 47/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 959us/step - accuracy: 0.9746 - loss: 0.0762 - val_accuracy: 0.7854 - val_loss: 1.1063\n",
      "Epoch 48/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9972 - loss: 0.0206 - val_accuracy: 0.7961 - val_loss: 1.4467\n",
      "Epoch 49/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9976 - loss: 0.0126 - val_accuracy: 0.8004 - val_loss: 1.6696\n",
      "Epoch 50/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9960 - loss: 0.0106 - val_accuracy: 0.7897 - val_loss: 1.7471\n",
      "Epoch 51/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9889 - loss: 0.0322 - val_accuracy: 0.7833 - val_loss: 1.3172\n",
      "Epoch 52/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 931us/step - accuracy: 0.9873 - loss: 0.0337 - val_accuracy: 0.7897 - val_loss: 1.3244\n",
      "Epoch 53/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 935us/step - accuracy: 0.9857 - loss: 0.0392 - val_accuracy: 0.8004 - val_loss: 1.2599\n",
      "Epoch 54/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9873 - loss: 0.0312 - val_accuracy: 0.7833 - val_loss: 1.4854\n",
      "Epoch 55/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9992 - loss: 0.0045 - val_accuracy: 0.7811 - val_loss: 1.6885\n",
      "Epoch 56/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 931us/step - accuracy: 0.9953 - loss: 0.0109 - val_accuracy: 0.7811 - val_loss: 1.5901\n",
      "Epoch 57/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 998us/step - accuracy: 0.9818 - loss: 0.0656 - val_accuracy: 0.7940 - val_loss: 1.1972\n",
      "Epoch 58/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 975us/step - accuracy: 0.9936 - loss: 0.0192 - val_accuracy: 0.7983 - val_loss: 1.4499\n",
      "Epoch 59/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 931us/step - accuracy: 0.9889 - loss: 0.0374 - val_accuracy: 0.7833 - val_loss: 1.2353\n",
      "Epoch 60/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9904 - loss: 0.0327 - val_accuracy: 0.7790 - val_loss: 1.4266\n",
      "Epoch 61/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 918us/step - accuracy: 1.0000 - loss: 0.0035 - val_accuracy: 0.7747 - val_loss: 1.9654\n",
      "Epoch 62/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 861us/step - accuracy: 1.0000 - loss: 3.3569e-04 - val_accuracy: 0.7747 - val_loss: 2.3092\n",
      "Epoch 63/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 948us/step - accuracy: 1.0000 - loss: 1.6892e-04 - val_accuracy: 0.7704 - val_loss: 2.5314\n",
      "Epoch 64/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 7.2569e-05 - val_accuracy: 0.7768 - val_loss: 2.6577\n",
      "Epoch 65/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 7.2654e-05 - val_accuracy: 0.7725 - val_loss: 2.7705\n",
      "Epoch 66/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 3.9437e-05 - val_accuracy: 0.7725 - val_loss: 2.8538\n",
      "Epoch 67/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 4.1582e-05 - val_accuracy: 0.7704 - val_loss: 2.9289\n",
      "Epoch 68/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 1.8892e-05 - val_accuracy: 0.7725 - val_loss: 2.9817\n",
      "Epoch 69/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 1.8131e-05 - val_accuracy: 0.7704 - val_loss: 3.0400\n",
      "Epoch 70/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 938us/step - accuracy: 1.0000 - loss: 1.9904e-05 - val_accuracy: 0.7704 - val_loss: 3.0946\n",
      "Epoch 71/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 1.9827e-05 - val_accuracy: 0.7747 - val_loss: 3.1410\n",
      "Epoch 72/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 1.7470e-05 - val_accuracy: 0.7725 - val_loss: 3.1845\n",
      "Epoch 73/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 1.3605e-05 - val_accuracy: 0.7725 - val_loss: 3.2222\n",
      "Epoch 74/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 1.0218e-05 - val_accuracy: 0.7725 - val_loss: 3.2573\n",
      "Epoch 75/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 1.2152e-05 - val_accuracy: 0.7704 - val_loss: 3.2939\n",
      "Epoch 76/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 1.1728e-05 - val_accuracy: 0.7704 - val_loss: 3.3254\n",
      "Epoch 77/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 1.3472e-05 - val_accuracy: 0.7704 - val_loss: 3.3560\n",
      "Epoch 78/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 8.8660e-06 - val_accuracy: 0.7704 - val_loss: 3.3849\n",
      "Epoch 79/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 963us/step - accuracy: 1.0000 - loss: 5.7242e-06 - val_accuracy: 0.7704 - val_loss: 3.4115\n",
      "Epoch 80/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 877us/step - accuracy: 1.0000 - loss: 6.4141e-06 - val_accuracy: 0.7704 - val_loss: 3.4389\n",
      "Epoch 81/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 6.1472e-06 - val_accuracy: 0.7682 - val_loss: 3.4642\n",
      "Epoch 82/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 3.6581e-06 - val_accuracy: 0.7682 - val_loss: 3.4863\n",
      "Epoch 83/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 4.8500e-06 - val_accuracy: 0.7704 - val_loss: 3.5141\n",
      "Epoch 84/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 921us/step - accuracy: 1.0000 - loss: 6.0898e-06 - val_accuracy: 0.7704 - val_loss: 3.5375\n",
      "Epoch 85/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 4.9514e-06 - val_accuracy: 0.7704 - val_loss: 3.5587\n",
      "Epoch 86/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 991us/step - accuracy: 1.0000 - loss: 7.3356e-06 - val_accuracy: 0.7704 - val_loss: 3.5808\n",
      "Epoch 87/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 899us/step - accuracy: 1.0000 - loss: 4.5561e-06 - val_accuracy: 0.7704 - val_loss: 3.6012\n",
      "Epoch 88/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 4.4042e-06 - val_accuracy: 0.7704 - val_loss: 3.6212\n",
      "Epoch 89/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 960us/step - accuracy: 1.0000 - loss: 4.5633e-06 - val_accuracy: 0.7704 - val_loss: 3.6406\n",
      "Epoch 90/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 926us/step - accuracy: 1.0000 - loss: 2.7448e-06 - val_accuracy: 0.7704 - val_loss: 3.6579\n",
      "Epoch 91/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 2ms/step - accuracy: 1.0000 - loss: 3.4195e-06 - val_accuracy: 0.7704 - val_loss: 3.6789\n",
      "Epoch 92/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 3.9085e-06 - val_accuracy: 0.7704 - val_loss: 3.6972\n",
      "Epoch 93/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 3.2672e-06 - val_accuracy: 0.7704 - val_loss: 3.7148\n",
      "Epoch 94/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 3.3254e-06 - val_accuracy: 0.7704 - val_loss: 3.7319\n",
      "Epoch 95/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 930us/step - accuracy: 1.0000 - loss: 2.4199e-06 - val_accuracy: 0.7704 - val_loss: 3.7486\n",
      "Epoch 96/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 2.1130e-06 - val_accuracy: 0.7704 - val_loss: 3.7652\n",
      "Epoch 97/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 2.6757e-06 - val_accuracy: 0.7704 - val_loss: 3.7820\n",
      "Epoch 98/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 1.7106e-06 - val_accuracy: 0.7704 - val_loss: 3.7969\n",
      "Epoch 99/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 2.0971e-06 - val_accuracy: 0.7704 - val_loss: 3.8135\n",
      "Epoch 100/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 1.0000 - loss: 1.5652e-06 - val_accuracy: 0.7725 - val_loss: 3.8275\n"
     ]
    }
   ],
   "source": [
    "# Define network\n",
    "number_features = X_train_sc.shape[1]\n",
    "model = make_net(number_features)\n",
    "\n",
    "### Train model (and store training info in history)\n",
    "history = model.fit(X_train_sc,\n",
    "                    y_train,\n",
    "                    epochs=100,\n",
    "                    batch_size=32,\n",
    "                    validation_data=(X_test_sc, y_test),\n",
    "                    verbose=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Calculate training and test set accuracy of your model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step\n",
      "\u001b[1m15/15\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 549us/step\n",
      "Training accuracy 1.000\n",
      "Test accuracy 0.773\n"
     ]
    }
   ],
   "source": [
    "calculate_accuracy(model, X_train_sc, X_test_sc, y_train, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Plot training and test set accuracy over time during fitting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\dan\\AppData\\Local\\Temp\\ipykernel_31272\\832266572.py:17: UserWarning: FigureCanvasAgg is non-interactive, and thus cannot be shown\n",
      "  fig.show()\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "history_dict = history.history\n",
    "history_dict.keys()\n",
    "\n",
    "plot_training(history.history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Save this baseline version of your model, so you can access it again later if you need it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save model\n",
    "model.save('baseline_stroke_dan.keras')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now try different things in your model to improve test accuracy.  You might consider :\n",
    "- Reducing overfitting if overfitting is a problem.\n",
    "- Changing the number of hidden layers\n",
    "- Changing the number of hidden neurons\n",
    "- Changing batch size\n",
    "- Changing dropout rate\n",
    "- Changing the learning rate\n",
    "- Changing the train / test split\n",
    "- Trying stratified k-fold validation\n",
    "- Dropping features\n",
    "\n",
    "or more!\n",
    "\n",
    "Tip : keep your analysis above as your base case.  Then below, just use the functions you've built to rebuild and retrain models with different parameters (or run altered versions of other cells below).  Don't forget, you need to build and train again before you get new outputs.\n",
    "\n",
    "Add comments to your code to explain what you've changed, and change things a bit at a time (don't change everything all at once!)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.6107 - loss: 0.6270 - val_accuracy: 0.8155 - val_loss: 0.4071\n",
      "Epoch 2/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7961 - loss: 0.4391 - val_accuracy: 0.8133 - val_loss: 0.3966\n",
      "Epoch 3/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8156 - loss: 0.4099 - val_accuracy: 0.8219 - val_loss: 0.3931\n",
      "Epoch 4/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8115 - loss: 0.4034 - val_accuracy: 0.7790 - val_loss: 0.5224\n",
      "Epoch 5/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7804 - loss: 0.4493 - val_accuracy: 0.8240 - val_loss: 0.3890\n",
      "Epoch 6/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8438 - loss: 0.3663 - val_accuracy: 0.8197 - val_loss: 0.3900\n",
      "Epoch 7/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8591 - loss: 0.3349 - val_accuracy: 0.8112 - val_loss: 0.4120\n",
      "Epoch 8/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8278 - loss: 0.3703 - val_accuracy: 0.8112 - val_loss: 0.4029\n",
      "Epoch 9/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8594 - loss: 0.3229 - val_accuracy: 0.7876 - val_loss: 0.4435\n",
      "Epoch 10/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8776 - loss: 0.3051 - val_accuracy: 0.8026 - val_loss: 0.4627\n",
      "Epoch 11/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 991us/step - accuracy: 0.8698 - loss: 0.2934 - val_accuracy: 0.7554 - val_loss: 0.6602\n",
      "Epoch 12/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 967us/step - accuracy: 0.8655 - loss: 0.2906 - val_accuracy: 0.8197 - val_loss: 0.4412\n",
      "Epoch 13/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 991us/step - accuracy: 0.8786 - loss: 0.2782 - val_accuracy: 0.8112 - val_loss: 0.4730\n",
      "Epoch 14/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8936 - loss: 0.2298 - val_accuracy: 0.8069 - val_loss: 0.5321\n",
      "Epoch 15/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 997us/step - accuracy: 0.9035 - loss: 0.2295 - val_accuracy: 0.8176 - val_loss: 0.5017\n"
     ]
    }
   ],
   "source": [
    "\"\"\"Add EarlyStopping with 10 epoch patience, and restore best weights\"\"\"\n",
    "\n",
    "checkpoint_cb = keras.callbacks.ModelCheckpoint(\n",
    "    'model_checkpoint.keras', save_best_only=True)\n",
    "\n",
    "# Define early stopping callback\n",
    "early_stopping_cb = keras.callbacks.EarlyStopping(\n",
    "    patience=10, restore_best_weights=True)\n",
    "\n",
    "# Define network\n",
    "number_features = X_train_sc.shape[1]\n",
    "model = make_net(number_features)\n",
    "\n",
    "### Train model (and store training info in history)\n",
    "history = model.fit(X_train_sc,\n",
    "                    y_train,\n",
    "                    epochs=100,\n",
    "                    batch_size=32,\n",
    "                    validation_data=(X_test_sc, y_test),\n",
    "                    verbose=1,\n",
    "                    callbacks=[checkpoint_cb, early_stopping_cb])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 992us/step\n",
      "\u001b[1m15/15\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 529us/step\n",
      "Training accuracy 0.835\n",
      "Test accuracy 0.824\n"
     ]
    }
   ],
   "source": [
    "calculate_accuracy(model, X_train_sc, X_test_sc, y_train, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\dan\\AppData\\Local\\Temp\\ipykernel_31272\\832266572.py:17: UserWarning: FigureCanvasAgg is non-interactive, and thus cannot be shown\n",
      "  fig.show()\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_training(history.history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.6508 - loss: 0.6134 - val_accuracy: 0.8004 - val_loss: 0.4267\n",
      "Epoch 2/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7875 - loss: 0.4342 - val_accuracy: 0.8197 - val_loss: 0.3939\n",
      "Epoch 3/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8147 - loss: 0.4055 - val_accuracy: 0.8112 - val_loss: 0.3909\n",
      "Epoch 4/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8235 - loss: 0.3869 - val_accuracy: 0.8369 - val_loss: 0.3855\n",
      "Epoch 5/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8459 - loss: 0.3681 - val_accuracy: 0.8391 - val_loss: 0.3859\n",
      "Epoch 6/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8339 - loss: 0.3546 - val_accuracy: 0.8369 - val_loss: 0.3840\n",
      "Epoch 7/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8555 - loss: 0.3348 - val_accuracy: 0.8197 - val_loss: 0.4127\n",
      "Epoch 8/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 956us/step - accuracy: 0.8794 - loss: 0.3000 - val_accuracy: 0.8047 - val_loss: 0.4339\n",
      "Epoch 9/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 938us/step - accuracy: 0.8790 - loss: 0.2948 - val_accuracy: 0.8197 - val_loss: 0.4363\n",
      "Epoch 10/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8838 - loss: 0.2697 - val_accuracy: 0.8391 - val_loss: 0.4655\n",
      "Epoch 11/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 992us/step - accuracy: 0.8887 - loss: 0.2523 - val_accuracy: 0.8240 - val_loss: 0.5127\n",
      "Epoch 12/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 929us/step - accuracy: 0.8853 - loss: 0.2642 - val_accuracy: 0.8262 - val_loss: 0.4821\n",
      "Epoch 13/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 862us/step - accuracy: 0.9054 - loss: 0.2427 - val_accuracy: 0.8283 - val_loss: 0.5086\n",
      "Epoch 14/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9248 - loss: 0.1904 - val_accuracy: 0.8262 - val_loss: 0.5621\n",
      "Epoch 15/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.9099 - loss: 0.2048 - val_accuracy: 0.8155 - val_loss: 0.6487\n",
      "Epoch 16/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 916us/step - accuracy: 0.9298 - loss: 0.1772 - val_accuracy: 0.7833 - val_loss: 0.7619\n"
     ]
    }
   ],
   "source": [
    "\"\"\"Keep EarlyStopping with 10 epoch patience, and restore best weights.\n",
    "Reduce number of hidden layers to 3 (from 5)\n",
    "\"\"\"\n",
    "\n",
    "checkpoint_cb = keras.callbacks.ModelCheckpoint(\n",
    "    'model_checkpoint.keras', save_best_only=True)\n",
    "\n",
    "# Define early stopping callback\n",
    "early_stopping_cb = keras.callbacks.EarlyStopping(\n",
    "    patience=10, restore_best_weights=True)\n",
    "\n",
    "# Define network\n",
    "number_features = X_train_sc.shape[1]\n",
    "model = make_net(number_features, hidden_layers=3)\n",
    "\n",
    "### Train model (and store training info in history)\n",
    "history = model.fit(X_train_sc,\n",
    "                    y_train,\n",
    "                    epochs=100,\n",
    "                    batch_size=32,\n",
    "                    validation_data=(X_test_sc, y_test),\n",
    "                    verbose=1,\n",
    "                    callbacks=[checkpoint_cb, early_stopping_cb])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 868us/step\n",
      "\u001b[1m15/15\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 650us/step\n",
      "Training accuracy 0.866\n",
      "Test accuracy 0.837\n"
     ]
    }
   ],
   "source": [
    "calculate_accuracy(model, X_train_sc, X_test_sc, y_train, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\dan\\AppData\\Local\\Temp\\ipykernel_31272\\832266572.py:17: UserWarning: FigureCanvasAgg is non-interactive, and thus cannot be shown\n",
      "  fig.show()\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkgAAAHHCAYAAABEEKc/AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/H5lhTAAAACXBIWXMAAA9hAAAPYQGoP6dpAACEVElEQVR4nO3dd1xV9f8H8NdlXTYoyDIExVnOUHGlln7FkebIlVvTLDdpam79GaallpqWucqZiSO3kiMTR+DImRNcgGiCgMz7+f3xiYuXoYBcDvfyej4eR88994z3uVy47/uZKiGEABERERFpmSgdABEREVFxwwSJiIiIKAsmSERERERZMEEiIiIiyoIJEhEREVEWTJCIiIiIsmCCRERERJQFEyQiIiKiLJggEREREWXBBImoCPTv3x/e3t4FOnb69OlQqVSFG1Axc/v2bahUKqxevbpIr3v48GGoVCocPnxYuy2vPyt9xezt7Y3+/fsX6jmJKP+YIFGJplKp8rQ8/wFK9KqOHz+O6dOn48mTJ0qHQkS5MFM6ACIl/fzzzzqPf/rpJxw4cCDb9mrVqr3SdZYvXw6NRlOgYydPnowJEya80vUp717lZ5VXx48fx4wZM9C/f384OjrqPHf16lWYmPC7K5HSmCBRida7d2+dxydOnMCBAweybc8qMTER1tbWeb6Oubl5geIDADMzM5iZ8Ve1qLzKz6owqNVqRa9vKBISEmBjY6N0GGTE+DWF6CWaN2+O6tWrIzQ0FE2bNoW1tTU+//xzAMD27dvRrl07eHh4QK1Ww8fHB7NmzUJ6errOObK2a8lov/LVV1/hhx9+gI+PD9RqNerVq4fTp0/rHJtTGySVSoXhw4dj27ZtqF69OtRqNd544w3s3bs3W/yHDx9G3bp1YWlpCR8fH3z//fd5btf0xx9/oGvXrihXrhzUajU8PT0xZswYPHv2LNv92dra4t69e+jYsSNsbW1RpkwZjB07Nttr8eTJE/Tv3x8ODg5wdHREv3798lTV9Ndff0GlUmHNmjXZntu3bx9UKhV27twJAAgPD8cnn3yCKlWqwMrKCk5OTujatStu37790uvk1AYprzGfP38e/fv3R4UKFWBpaQk3NzcMHDgQjx490u4zffp0jBs3DgBQvnx5bTVuRmw5tUG6efMmunbtitKlS8Pa2hoNGjTArl27dPbJaE/1yy+/YPbs2XjttddgaWmJFi1a4Pr16y+97/y8Zk+ePMGYMWPg7e0NtVqN1157DX379kVMTIx2n6SkJEyfPh2VK1eGpaUl3N3d0blzZ9y4cUMn3qzV1zm17cp4f924cQNt27aFnZ0devXqBSDv71EAuHLlCrp164YyZcrAysoKVapUwaRJkwAAhw4dgkqlwtatW7Mdt379eqhUKoSEhLz0dSTjwa+lRHnw6NEjtGnTBj169EDv3r3h6uoKAFi9ejVsbW0REBAAW1tb/P7775g6dSri4uIwb968l553/fr1ePr0KT766COoVCrMnTsXnTt3xs2bN19aknHs2DEEBQXhk08+gZ2dHb799lt06dIFERERcHJyAgCcOXMGrVu3hru7O2bMmIH09HTMnDkTZcqUydN9b968GYmJifj444/h5OSEU6dOYdGiRbh79y42b96ss296ejr8/f3h5+eHr776CgcPHsTXX38NHx8ffPzxxwAAIQTee+89HDt2DEOHDkW1atWwdetW9OvX76Wx1K1bFxUqVMAvv/ySbf9NmzahVKlS8Pf3BwCcPn0ax48fR48ePfDaa6/h9u3bWLp0KZo3b45Lly7lq/QvPzEfOHAAN2/exIABA+Dm5oaLFy/ihx9+wMWLF3HixAmoVCp07twZ//zzDzZs2IAFCxbA2dkZAHL9mURFRaFRo0ZITEzEyJEj4eTkhDVr1qBDhw749ddf0alTJ53958yZAxMTE4wdOxaxsbGYO3cuevXqhZMnT77wPvP6msXHx+Ott97C5cuXMXDgQLz55puIiYnBjh07cPfuXTg7OyM9PR3vvvsugoOD0aNHD4waNQpPnz7FgQMHcOHCBfj4+OT59c+QlpYGf39/NGnSBF999ZU2nry+R8+fP4+33noL5ubmGDJkCLy9vXHjxg389ttvmD17Npo3bw5PT0+sW7cu22u6bt06+Pj4oGHDhvmOmwyYICKtYcOGiay/Fs2aNRMAxLJly7Ltn5iYmG3bRx99JKytrUVSUpJ2W79+/YSXl5f28a1btwQA4eTkJB4/fqzdvn37dgFA/Pbbb9pt06ZNyxYTAGFhYSGuX7+u3Xbu3DkBQCxatEi7rX379sLa2lrcu3dPu+3atWvCzMws2zlzktP9BQYGCpVKJcLDw3XuD4CYOXOmzr516tQRvr6+2sfbtm0TAMTcuXO129LS0sRbb70lAIhVq1a9MJ6JEycKc3NzndcsOTlZODo6ioEDB74w7pCQEAFA/PTTT9pthw4dEgDEoUOHdO7l+Z9VfmLO6bobNmwQAMTRo0e12+bNmycAiFu3bmXb38vLS/Tr10/7ePTo0QKA+OOPP7Tbnj59KsqXLy+8vb1Fenq6zr1Uq1ZNJCcna/f95ptvBADx999/Z7vW8/L6mk2dOlUAEEFBQdn212g0QgghVq5cKQCI+fPn57pPTq+9EJm/G8+/rhnvrwkTJuQp7pzeo02bNhV2dnY6256PRwj5/lKr1eLJkyfabdHR0cLMzExMmzYt23XIuLGKjSgP1Go1BgwYkG27lZWVdv3p06eIiYnBW2+9hcTERFy5cuWl5+3evTtKlSqlffzWW28BkFUqL9OyZUudb+I1a9aEvb299tj09HQcPHgQHTt2hIeHh3a/ihUrok2bNi89P6B7fwkJCYiJiUGjRo0ghMCZM2ey7T906FCdx2+99ZbOvezevRtmZmbaEiUAMDU1xYgRI/IUT/fu3ZGamoqgoCDttv379+PJkyfo3r17jnGnpqbi0aNHqFixIhwdHREWFpanaxUk5uevm5SUhJiYGDRo0AAA8n3d569fv359NGnSRLvN1tYWQ4YMwe3bt3Hp0iWd/QcMGAALCwvt47y+p/L6mm3ZsgW1atXKVsoCQFttu2XLFjg7O+f4Gr3KkBXP/wxyiju39+jDhw9x9OhRDBw4EOXKlcs1nr59+yI5ORm//vqrdtumTZuQlpb20naJZHyYIBHlQdmyZXU+dDJcvHgRnTp1goODA+zt7VGmTBntH9LY2NiXnjfrH+uMZOnff//N97EZx2ccGx0djWfPnqFixYrZ9stpW04iIiLQv39/lC5dWtuuqFmzZgCy35+lpWW2aqLn4wFkOxd3d3fY2trq7FelSpU8xVOrVi1UrVoVmzZt0m7btGkTnJ2d8c4772i3PXv2DFOnToWnpyfUajWcnZ1RpkwZPHnyJE8/l+flJ+bHjx9j1KhRcHV1hZWVFcqUKYPy5csDyNv7Ibfr53StjJ6V4eHhOtsL+p7K62t248YNVK9e/YXnunHjBqpUqVKonQvMzMzw2muvZduel/doRnL4srirVq2KevXqYd26ddpt69atQ4MGDfL8O0PGg22QiPLg+W+pGZ48eYJmzZrB3t4eM2fOhI+PDywtLREWFobx48fnqau4qalpjtuFEHo9Ni/S09Pxv//9D48fP8b48eNRtWpV2NjY4N69e+jfv3+2+8stnsLWvXt3zJ49GzExMbCzs8OOHTvQs2dPnQ/jESNGYNWqVRg9ejQaNmwIBwcHqFQq9OjRQ69d+Lt164bjx49j3LhxqF27NmxtbaHRaNC6dWu9Dx2QoaDvi6J+zXIrScraqD+DWq3ONvxBft+jedG3b1+MGjUKd+/eRXJyMk6cOIHFixfn+zxk+JggERXQ4cOH8ejRIwQFBaFp06ba7bdu3VIwqkwuLi6wtLTMsQdTXno1/f333/jnn3+wZs0a9O3bV7v9wIEDBY7Jy8sLwcHBiI+P1ymRuXr1ap7P0b17d8yYMQNbtmyBq6sr4uLi0KNHD519fv31V/Tr1w9ff/21dltSUlKBBmbMa8z//vsvgoODMWPGDEydOlW7/dq1a9nOmZ9qJi8vrxxfn4wqXC8vrzyf60Xy+pr5+PjgwoULLzyXj48PTp48idTU1Fw7G2SUbGU9f9YSsRfJ63u0QoUKAPDSuAGgR48eCAgIwIYNG/Ds2TOYm5vrVN9SycEqNqICyvim/vw385SUFHz33XdKhaTD1NQULVu2xLZt23D//n3t9uvXr2PPnj15Oh7QvT8hBL755psCx9S2bVukpaVh6dKl2m3p6elYtGhRns9RrVo11KhRA5s2bcKmTZvg7u6uk6BmxJ61xGTRokW5lk4URsw5vV4AsHDhwmznzBi/Jy8JW9u2bXHq1CmdLuYJCQn44Ycf4O3tjddffz2vt/JCeX3NunTpgnPnzuXYHT7j+C5duiAmJibHkpeMfby8vGBqaoqjR4/qPJ+f35+8vkfLlCmDpk2bYuXKlYiIiMgxngzOzs5o06YN1q5di3Xr1qF169banoZUsrAEiaiAGjVqhFKlSqFfv34YOXIkVCoVfv7550Kr4ioM06dPx/79+9G4cWN8/PHHSE9Px+LFi1G9enWcPXv2hcdWrVoVPj4+GDt2LO7duwd7e3ts2bIlT+2jctO+fXs0btwYEyZMwO3bt/H6668jKCgo3+1zunfvjqlTp8LS0hKDBg3KVvXy7rvv4ueff4aDgwNef/11hISE4ODBg9rhD/QRs729PZo2bYq5c+ciNTUVZcuWxf79+3MsUfT19QUATJo0CT169IC5uTnat2+f48CHEyZMwIYNG9CmTRuMHDkSpUuXxpo1a3Dr1i1s2bKl0EbdzutrNm7cOPz666/o2rUrBg4cCF9fXzx+/Bg7duzAsmXLUKtWLfTt2xc//fQTAgICcOrUKbz11ltISEjAwYMH8cknn+C9996Dg4MDunbtikWLFkGlUsHHxwc7d+5EdHR0nmPOz3v022+/RZMmTfDmm29iyJAhKF++PG7fvo1du3Zl+13o27cv3n//fQDArFmz8v9iklFggkRUQE5OTti5cyc+/fRTTJ48GaVKlULv3r3RokUL7Xg8SvP19cWePXswduxYTJkyBZ6enpg5cyYuX7780l525ubm+O233zBy5EgEBgbC0tISnTp1wvDhw1GrVq0CxWNiYoIdO3Zg9OjRWLt2LVQqFTp06ICvv/4aderUyfN5unfvjsmTJyMxMTHH6o9vvvkGpqamWLduHZKSktC4cWMcPHiwQD+X/MS8fv16jBgxAkuWLIEQAq1atcKePXt0ehECQL169TBr1iwsW7YMe/fuhUajwa1bt3JMkFxdXXH8+HGMHz8eixYtQlJSEmrWrInffvsN7dq1y/f95Cavr5mtrS3++OMPTJs2DVu3bsWaNWvg4uKCFi1aaBtRm5qaYvfu3Zg9ezbWr1+PLVu2wMnJCU2aNEGNGjW051q0aBFSU1OxbNkyqNVqdOvWDfPmzXtpY+oM+XmP1qpVCydOnMCUKVOwdOlSJCUlwcvLC926dct23vbt26NUqVLQaDTo0KFDfl9KMhIqUZy+7hJRkejYsSMuXryYY/sYopIuLS0NHh4eaN++PVasWKF0OKQQtkEiMnJZp1y4du0adu/ejebNmysTEFExt23bNjx8+FCn4TeVPCxBIjJy7u7u2vnBwsPDsXTpUiQnJ+PMmTOoVKmS0uERFRsnT57E+fPnMWvWLDg7Oxd4cE8yDmyDRGTkWrdujQ0bNiAyMhJqtRoNGzbEF198weSIKIulS5di7dq1qF27ts5kuVQysQSJiIiIKAu2QSIiIiLKggkSERERURZsg1RAGo0G9+/fh52d3SvNTk1ERERFRwiBp0+fwsPD44UDrTJBKqD79+/D09NT6TCIiIioAO7cuaMd3DQnTJAKyM7ODoB8ge3t7RWOhoiIiPIiLi4Onp6e2s/x3DBBKqCMajV7e3smSERERAbmZc1j2EibiIiIKAsmSERERERZMEEiIiIiyoJtkPQsPT0dqampSodBBsLc3BympqZKh0FEVOIxQdITIQQiIyPx5MkTpUMhA+Po6Ag3NzeOr0VEpCAmSHqSkRy5uLjA2tqaH3b0UkIIJCYmIjo6GgDg7u6ucERERCUXEyQ9SE9P1yZHTk5OSodDBsTKygoAEB0dDRcXF1a3EREphI209SCjzZG1tbXCkZAhynjfsO0aEZFymCDpEavVqCD4viEiUh4TJCIiIqIsFE+QlixZAm9vb1haWsLPzw+nTp3Kdd/U1FTMnDkTPj4+sLS0RK1atbB3716dfQIDA1GvXj3Y2dnBxcUFHTt2xNWrV3X2ad68OVQqlc4ydOhQvdxfSeft7Y2FCxfmef/Dhw9DpVKx9x8RESlK0QRp06ZNCAgIwLRp0xAWFoZatWrB399f24snq8mTJ+P777/HokWLcOnSJQwdOhSdOnXCmTNntPscOXIEw4YNw4kTJ3DgwAGkpqaiVatWSEhI0DnX4MGD8eDBA+0yd+5cvd5rcZc1Ycy6TJ8+vUDnPX36NIYMGZLn/Rs1aoQHDx7AwcGhQNcjIiIqDCohhFDq4n5+fqhXrx4WL14MANBoNPD09MSIESMwYcKEbPt7eHhg0qRJGDZsmHZbly5dYGVlhbVr1+Z4jYcPH8LFxQVHjhxB06ZNAcgSpNq1a+erZCOruLg4ODg4IDY2NttktUlJSbh16xbKly8PS0vLAl+jKEVGRmrXN23ahKlTp+qUvNna2sLW1haA7I6enp4OMzN2gtQHQ3z/EBEVpqQk4OhRoFWrwj/3iz6/n6dYCVJKSgpCQ0PRsmXLzGBMTNCyZUuEhITkeExycnK2DwwrKyscO3Ys1+vExsYCAEqXLq2zfd26dXB2dkb16tUxceJEJCYmFvRWjIKbm5t2cXBwgEql0j6+cuUK7OzssGfPHvj6+kKtVuPYsWO4ceMG3nvvPbi6usLW1hb16tXDwYMHdc6btYpNpVLhxx9/RKdOnWBtbY1KlSphx44d2uezVrGtXr0ajo6O2LdvH6pVqwZbW1u0bt0aDx480B6TlpaGkSNHwtHREU5OThg/fjz69euHjh075nq/jx49Qs+ePVG2bFlYW1ujRo0a2LBhg84+Go0Gc+fORcWKFaFWq1GuXDnMnj1b+/zdu3fRs2dPlC5dGjY2Nqhbty5OnjxZgFefiIgy3LgBNGoEtG0L/PGHcnEoliDFxMQgPT0drq6uOttdXV11SjOe5+/vj/nz5+PatWvQaDQ4cOAAgoKCdD4sn6fRaDB69Gg0btwY1atX127/4IMPsHbtWhw6dAgTJ07Ezz//jN69e78w3uTkZMTFxekseSUEkJCgzFKY5YMTJkzAnDlzcPnyZdSsWRPx8fFo27YtgoODcebMGbRu3Rrt27dHRETEC88zY8YMdOvWDefPn0fbtm3Rq1cvPH78ONf9ExMT8dVXX+Hnn3/G0aNHERERgbFjx2qf//LLL7Fu3TqsWrUKf/75J+Li4rBt27YXxpCUlARfX1/s2rULFy5cwJAhQ9CnTx+dNnATJ07EnDlzMGXKFFy6dAnr16/Xvl/j4+PRrFkz3Lt3Dzt27MC5c+fw2WefQaPR5OGVJCKinAQFAW++CZw5A5QqBaSkKBiMUMi9e/cEAHH8+HGd7ePGjRP169fP8Zjo6Gjx3nvvCRMTE2FqaioqV64sPvnkE2FpaZnj/kOHDhVeXl7izp07L4wlODhYABDXr1/PdZ9p06YJANmW2NjYbPs+e/ZMXLp0STx79kwIIUR8vBAyVSn6JT7+hbeeo1WrVgkHBwft40OHDgkAYtu2bS899o033hCLFi3SPvby8hILFizQPgYgJk+erH0cHx8vAIg9e/boXOvff//VxpL1Z7NkyRLh6uqqfezq6irmzZunfZyWlibKlSsn3nvvvbzeshBCiHbt2olPP/1UCCFEXFycUKvVYvny5Tnu+/333ws7Ozvx6NGjfF0jL7K+f4iIjF1yshCjR2d+djVuLMRLProLLDY2NtfP7+cpVoLk7OwMU1NTREVF6WyPioqCm5tbjseUKVMG27ZtQ0JCAsLDw3HlyhXY2tqiQoUK2fYdPnw4du7ciUOHDuG11157YSx+fn4AgOvXr+e6z8SJExEbG6td7ty587JbNDp169bVeRwfH4+xY8eiWrVqcHR0hK2tLS5fvvzSEqSaNWtq121sbGBvb59rw3xADpzo4+Ojfezu7q7dPzY2FlFRUahfv772eVNTU/j6+r4whvT0dMyaNQs1atRA6dKlYWtri3379mljv3z5MpKTk9GiRYscjz979izq1KmTreqWiIjyJyICaNoUyGiNMW4ccOgQ8JKPbr1TrJWthYUFfH19ERwcrG0rotFoEBwcjOHDh7/wWEtLS5QtWxapqanYsmULunXrpn1OCIERI0Zg69atOHz4MMqXL//SWM6ePQvgxXNfqdVqqNXql99YDqytgfj4Ah36ygpzMG8bGxudx2PHjsWBAwfw1VdfoWLFirCyssL777+PlJeUiZqbm+s8VqlUL6yayml/8Yp1h/PmzcM333yDhQsXokaNGrCxscHo0aO1sWdM+ZGblz1PREQvt2sX0Lcv8Pgx4OgIrFkDdOigdFSSot2QAgIC0K9fP9StWxf169fHwoULkZCQgAEDBgAA+vbti7JlyyIwMBAAcPLkSdy7dw+1a9fGvXv3MH36dGg0Gnz22Wfacw4bNgzr16/H9u3bYWdnp23P5ODgACsrK9y4cQPr169H27Zt4eTkhPPnz2PMmDFo2rSpTslGYVKpgCy5hVH4888/0b9/f3Tq1AmALFG6fft2kcbg4OAAV1dXnD59WttLMT09HWFhYahdu3aux/3555947733tG3PNBoN/vnnH7z++usAgEqVKsHKygrBwcH48MMPsx1fs2ZN/Pjjj3j8+DFLkYiI8iktDZgyBZgzRz6uWxfYvBnw9lY0LB2KJkjdu3fHw4cPMXXqVERGRqJ27drYu3evtiFsREQETEwyawGTkpIwefJk3Lx5E7a2tmjbti1+/vlnODo6avdZunQpANmV/3mrVq1C//79YWFhgYMHD2qTMU9PT3Tp0gWTJ0/W+/0am0qVKiEoKAjt27eHSqXClClTFGmkPGLECAQGBqJixYqoWrUqFi1ahH///feFU3ZUqlQJv/76K44fP45SpUph/vz5iIqK0iZIlpaWGD9+PD777DNYWFigcePGePjwIS5evIhBgwahZ8+e+OKLL9CxY0cEBgbC3d0dZ86cgYeHBxo2bFhUt05EZHDu3wd69pTd+AFgxAhg3jyggJU0eqP4QDbDhw/PtUrt8OHDOo+bNWuGS5cuvfB8L6t68fT0xJEjR/IVI+Vs/vz5GDhwIBo1agRnZ2eMHz8+X737Csv48eMRGRmJvn37wtTUFEOGDIG/vz9MTU1zPSYj0fb394e1tTWGDBmCjh07aoeFAIApU6bAzMwMU6dOxf379+Hu7q4dcd3CwgL79+/Hp59+irZt2yItLQ2vv/46lixZovf7JSIyVMHBwAcfANHRgJ0d8OOPwHOtZIoVRQeKNGTGNlCkMdFoNKhWrRq6deuGWbNmKR1OvvH9Q0TGJj0dmD0bmD5d9lOrWVNWqVWuXPSx5HWgSMVLkIheVXh4OPbv349mzZohOTkZixcvxq1bt/DBBx8oHRoRUYkXHQ307g0cOCAff/gh8O23QHHv68IEiQyeiYkJVq9ejbFjx0IIgerVq+PgwYOoVq2a0qEREZVox44B3bvLdkfW1sDSpbLXmiFggkQGz9PTE3/++afSYRAR0X80GuDrr4GJE2X1WtWqwK+/Am+8oXRkeccEiYiIiArN48dA//7Ab7/Jxx98AHz/PfDffOcGgwkSERERFYpTp2SvtPBw2W3/22+BwYPleICGRrGpRoiIiMg4CAEsWgQ0aSKTIx8fICQEGDLEMJMjgCVIRERE9Ari4mTPtM2b5ePOnYGVKwEHB2XjelVMkIiIiIoxjQY4fx7491+gVi2gOM1udO4c8P77wPXrgJkZ8NVXwMiRhltq9DwmSERERMXMw4fA/v3Avn1yiY7OfK58ecDXV3cp6qRJCGDFCjlNSFIS4OkJ/PIL0KBB0cahT0yQiIiIFJaaCpw4IZOhvXuBsDCZhGSwsQFcXIBbtzKXX3/NfN7bW074WhRJU0IC8PHHwM8/y8dt2wI//QQ4OennekphgkQA8MKJXQFg2rRpmD59eoHPvXXrVnTs2LFAxxMRGaPw8MwSooMHZVue59WqBbRuDfj7A40bAxYWspotLAwIDc1cbtwAbt+WS9akKWtJ06smMZcvyyq1S5cAU1M5fci4cYCJEXb5YoJEAIAHDx5o1zdt2oSpU6fi6tWr2m22hjaARSFISUmBhYWF0mEQkZF49gw4ciSzlOjKFd3nnZyAVq1kQtSqFeDunv0cpUoBLVrIJcO//wJnzshk6a+/sidNW7Zk7vsqSdO6dbJXWmKijG3jRqBp03y+CIZEUIHExsYKACI2Njbbc8+ePROXLl0Sz549UyCyV7dq1Srh4OCgs2358uWiatWqQq1WiypVqoglS5Zon0tOThbDhg0Tbm5uQq1Wi3LlyokvvvhCCCGEl5eXAKBdvLy8cr3uZ599JipVqiSsrKxE+fLlxeTJk0VKSorOPjt27BB169YVarVaODk5iY4dO2qfS0pKEp999pl47bXXhIWFhfDx8RE//vhjrve0detW8fyvwLRp00StWrXE8uXLhbe3t1CpVEIIIfbs2SMaN24sHBwcROnSpUW7du3E9evXdc51584d0aNHD1GqVClhbW0tfH19xYkTJ8StW7eESqUSp0+f1tl/wYIFoly5ciI9PT3b62Do7x8ikjQaIS5eFGL+fCFatRLC0lIIWXEmFxMTIRo3FmLmTCFOnRIiLa3wrv3vv0IEBwsxd64Q3bsLUbGi7rWfX7y8hOjcWYjZs4XYt0+ImBjdcz17JsSQIZn7t2ghRGRk4cVa1F70+f08liAVBSFkyq0Ea+tX7k6wbt06TJ06FYsXL0adOnVw5swZDB48GDY2NujXrx++/fZb7NixA7/88gvKlSuHO3fu4M6dOwCA06dPw8XFBatWrULr1q1hamqa63Xs7OywevVqeHh44O+//8bgwYNhZ2eHzz77DACwa9cudOrUCZMmTcJPP/2ElJQU7N69W3t83759ERISgm+//Ra1atXCrVu3EBMTk697vX79OrZs2YKgoCBtrAkJCQgICEDNmjURHx+PqVOnolOnTjh79ixMTEwQHx+PZs2aoWzZstixYwfc3NwQFhYGjUYDb29vtGzZEqtWrULdunW111m1ahX69+8PE2MslyYqwZ48AYKDZQnRvn3Af38KtTw9ZQmRv78sBSpVSj9xODoC77wjl+djyyhpyliuXZNVfeHhQFBQ5r5eXrJ06c03ZbXd2bPyo2TqVGDKFFm9ZvSKKGEzOvkqQYqPzz111/cSH5/ve8ta2uLj4yPWr1+vs8+sWbNEw4YNhRBCjBgxQrzzzjtCo9HkeD4AYuvWrfmOY968ecLX11f7uGHDhqJXr1457nv16lUBQBw4cCDH5/NagmRubi6io6NfGNfDhw8FAPH3338LIYT4/vvvhZ2dnXj06FGO+2/atEmUKlVKJCUlCSGECA0NFSqVSty6dSvH/VmCRGQ40tNl6c/MmbI0yNRU90+wWi1Lj+bPl6VJufyZVMyTJ0L8/rsQ8+YJ0aOHEJUq5fxRUqaMEPv3Kx1t4WAJEhWKhIQE3LhxA4MGDcLgwYO129PS0uDw3yhg/fv3x//+9z9UqVIFrVu3xrvvvotWrVrl+1qbNm3Ct99+ixs3biA+Ph5paWmwt7fXPn/27FmdGJ539uxZmJqaolmzZvm+7vO8vLxQpkwZnW3Xrl3D1KlTcfLkScTExECj0QAAIiIiUL16dZw9exZ16tRB6Vy6jHTs2BHDhg3D1q1b0aNHD6xevRpvv/02vL29XylWIlJGZGRm4+r9+4FHj3Sfr1pVlhC1bi3b6FhbKxNnXjg4AG+/LZcMsbG6JU3W1sCMGUDZssrFqQQmSEXB2hqIj1fu2q8g/r+4ly9fDj8/P53nMqqg3nzzTdy6dQt79uzBwYMH0a1bN7Rs2RK/Pt+d4iVCQkLQq1cvzJgxA/7+/nBwcMDGjRvx9ddfa/exsrLK9fgXPQcAJiYmEM/3mQWQmpqabT8bG5ts29q3bw8vLy8sX74cHh4e0Gg0qF69OlJSUvJ0bQsLC/Tt2xerVq1C586dsX79enzzzTcvPIaIip/0dKBXL2DTJt3t9vayuiyjx5mXlzLxFRYHB6B5c7mUZEyQioJKJQexMECurq7w8PDAzZs30atXr1z3s7e3R/fu3dG9e3e8//77aN26NR4/fozSpUvD3Nwc6enpL7zO8ePH4eXlhUmTJmm3hYeH6+xTs2ZNBAcHY8CAAdmOr1GjBjQaDY4cOYKWLVtme75MmTJ4+vQpEhIStEnQ2bNnXxgTADx69AhXr17F8uXL8dZbbwEAjh07li2uH3/8UXu/Ofnwww9RvXp1fPfdd0hLS0Pnzp1fem0iKl6WLctMjnx9MxOiBg0Ac3NlY6PCxwSJXmrGjBkYOXIkHBwc0Lp1ayQnJ+Ovv/7Cv//+i4CAAMyfPx/u7u6oU6cOTExMsHnzZri5ucHR0REA4O3tjeDgYDRu3BhqtRqlcmiVWKlSJURERGDjxo2oV68edu3aha1bt+rsM23aNLRo0QI+Pj7o0aMH0tLSsHv3bowfPx7e3t7o168fBg4cqG2kHR4ejujoaHTr1g1+fn6wtrbG559/jpEjR+LkyZNYvXr1S++9VKlScHJywg8//AB3d3dERERgwoQJOvv07NkTX3zxBTp27IjAwEC4u7vjzJkz8PDwQMOGDQEA1apVQ4MGDTB+/HgMHDjwpaVORFS83L8PTJwo1xcvBoYNUzYeKgJF0yTK+JS0bv7r1q0TtWvXFhYWFqJUqVKiadOmIigoSAghxA8//CBq164tbGxshL29vWjRooUICwvTHrtjxw5RsWJFYWZm9sJu/uPGjRNOTk7C1tZWdO/eXSxYsCBbHFu2bNHG4ezsLDp37qx97tmzZ2LMmDHC3d1dWFhYiIoVK4qVK1dqn9+6dauoWLGisLKyEu+++6744Ycfcuzmn9WBAwdEtWrVhFqtFjVr1hSHDx/O1vD89u3bokuXLsLe3l5YW1uLunXripMnT+qcZ8WKFQKAOHXqVK6vQcZ9GPL7h8gYvf++bKzs51e43fGp6OW1kbZKiCwNMyhP4uLi4ODggNjYWJ2GxACQlJSEW7duoXz58rC0tFQoQipuZs2ahc2bN+P8+fMv3I/vH6LiZedOoH172bU9LAyoWVPpiOhVvOjz+3kchIVIz+Lj43HhwgUsXrwYI0aMUDocIsqH+PjM6rRPP2VyVJIwQSLSs+HDh8PX1xfNmzfHwIEDlQ6HiPJh+nQgIkJO0TF1qtLRUFFiI20iPVu9enWeGoQTUfFy5gywcKFc/+47g+2MTAXEEiQiIqIs0tPlxKzp6UD37kCbNkpHREWNCZIesf07FQTfN0TK++474K+/5KCJGaVIVLIwQdID8/9GDEtUaoJaMmgZ7xtzjjxHpIi7d4HPP5frX34JuLkpGw8pg22Q9MDU1BSOjo6Ijo4GAFhbW0OlUikcFRV3QggkJiYiOjoajo6O2qlciKhojRwpe681bAjkMv0jlQBMkPTE7b+vHBlJElFeOTo6at8/RFS0tm8Htm4FzMyAH34ATFjPUmIxQdITlUoFd3d3uLi45DgpKlFOzM3NWXJEpJCnT4Hhw+X6uHFA9erKxkMK0/+g3i+2ePFi4eXlJdRqtahfv3626Rmel5KSImbMmCEqVKignfZhz549+T7ns2fPxCeffCJKly4tbGxsROfOnUVkZGS+4s7rUOVERGQYRo+W04lUqCBEQoLS0ZC+5PXzW9HCw02bNiEgIADTpk1DWFgYatWqBX9//1yrpSZPnozvv/8eixYtwqVLlzB06FB06tQJZ86cydc5x4wZg99++w2bN2/GkSNHcP/+fc6uTkRUgoWGAt9+K9eXLgWsrZWNh4qBIkrYclS/fn0xbNgw7eP09HTh4eEhAgMDc9zf3d1dLF68WGdb586dRa9evfJ8zidPnghzc3OxefNm7T6XL18WAERISEieY2cJEhGRcUhNFeLNN2Xp0QcfKB0N6VuxL0FKSUlBaGgoWrZsqd1mYmKCli1bIiQkJMdjkpOTs03eaWVlhWPHjuX5nKGhoUhNTdXZp2rVqihXrlyu1yUiIuO1eLGchNbREZg/X+loqLhQLEGKiYlBeno6XF1ddba7uroiMjIyx2P8/f0xf/58XLt2DRqNBgcOHEBQUBAePHiQ53NGRkbCwsICjo6Oeb4uIJOzuLg4nYWIiAzbnTvA5Mlyfe5cIMvHB5VgBtWB8ZtvvkGlSpVQtWpVWFhYYPjw4RgwYABMiqAfZmBgIBwcHLSLp6en3q9JRET6I4TstZaQADRuDAwapHREVJwoliA5OzvD1NQUUVFROtujoqJyHQOmTJky2LZtGxISEhAeHo4rV67A1tYWFSpUyPM53dzckJKSgidPnuT5ugAwceJExMbGapc7d+7k95aJiKgY2bYN2LEDMDfnmEeUnWJvBwsLC/j6+iI4OFi7TaPRIDg4GA0bNnzhsZaWlihbtizS0tKwZcsWvPfee3k+p6+vL8zNzXX2uXr1KiIiIl54XbVaDXt7e52FiIgMU1wcMGKEXP/sM+D115WNh4ofRQeKDAgIQL9+/VC3bl3Ur18fCxcuREJCAgYMGAAA6Nu3L8qWLYvAwEAAwMmTJ3Hv3j3Url0b9+7dw/Tp06HRaPDZZ5/l+ZwODg4YNGgQAgICULp0adjb22PEiBFo2LAhGjRoUPQvAhERFbnJk4F794CKFYFJk5SOhoojRROk7t274+HDh5g6dSoiIyNRu3Zt7N27V9vIOiIiQqd9UVJSEiZPnoybN2/C1tYWbdu2xc8//6zT4Ppl5wSABQsWwMTEBF26dEFycjL8/f3x3XffFdl9ExGRck6flj3XADnmkZWVsvFQ8aQSQgilgzBEcXFxcHBwQGxsLKvbiIgMRFoaUK8ecPYs0Ls38PPPSkdERS2vn99skkZERCXGN9/I5Kh0aeDrr5WOhoozJkhERFQihIcDU6fK9XnzABcXZeOh4o0JEhERGT0hgGHDgMREoGlT4L9+O0S5YoJERERGb8sWYNcuOebRsmWASqV0RFTcMUEiIiKjFhsLjBwp1ydOBKpVUzYeMgxMkIiIyKh9/jnw4AFQqZJMkIjyggkSEREZrRMn5FhHgKxas7RUNh4yHEyQiIjIKKWmAh99JBto9+sHvPOO0hGRIWGCRERERmnBAuD8ecDJCfjqK6WjIUPDBImIiIzOrVvA9Oly/auvAGdnRcMhA8QEiYiIjIoQwCefAM+eAc2by+o1ovxigkREREbll1+AvXsBCwuOeUQFxwSJiIiMxpMnwKhRcv3zz4EqVRQNhwwYEyQiIjIaEycCUVEyMZowQeloyJAxQSIiIqNw/LisUgOA778H1Gpl4yHDxgSJiIgMXsaYR4CciLZZM2XjIcPHBImIiAze118DFy7I7vzz5ikdDRkDJkhERGTQbtwAZsyQ6/Pny4EhiV4VEyQiIjJYGWMeJSUBLVoAvXsrHREZCzOlAyAiIsOSlgbcvAlcvgxcuSLHGSpXDvDykoubG2BSRF+/N24E9u+XDbKXLuWYR1R4mCAREVGOEhOBq1dlIvT8cu2abBSdG3NzwNNTJkvPJ04Z656egKXlq8f3+DEwerRcnzwZqFTp1c9JlIEJEhFRCffoUfYk6MoVIDxcVmHlxMoKqFpVLiYmQESE3P/ePZk83bwpl9y4umZPnJ5PphwdX14aNGECEB0NVKsGfPZZgW+fKEdMkIiISgAhgDt3MpOf55Ohhw9zP87JSSYgVavK/zOWcuVyrkZLSwPu35fJUnh4ZuL0/HpiohzMMSoKOHUq5+va2mZPoJ5fv3kTWL5c7vv993JaEaLCpBIit+8H9CJxcXFwcHBAbGws7O3tlQ6HiAiALL25cSPnEqGEhNyP8/TUTYAyljJlCjc+IWSJVU6JU8b6ixK2rD78MDNRIsqLvH5+swSJiMiAPX0K7N4NbN8OnDkDXL8uS3FyYmYGVKyYPQmqUkWW2BQFlUqOVeTsDLz5Zs77JCbK0q7cEqi7d+U9vvYa8OWXRRM3lTxMkIiIDExMDPDbb0BQEHDgAJCcrPu8jY1ulVjGesWKsgF1cWdtLZO23CaaTU8HHjyQ7ZSKKrGjkocJEhGRAbh7F9i2TSZFR44AGk3mcxUrAl26AG+/LROh114rum72SjA1lfdIpE9MkIiIiqlr12RCFBSUvTFz7dpAp05A587AG29w/B+iwsYEiYiomBACOHdOJkRbt8q5xTKoVEDDhjIh6tQJqFBBuTiJSgImSERECtJogBMnMkuKbt3KfM7MTFabde4MvPce4O6uXJxEJQ0TJCKiIpaaChw+LBOibduAyMjM5ywtgdatZVL07rtAqVJKRUlUsinejG/JkiXw9vaGpaUl/Pz8cCq3UcP+s3DhQlSpUgVWVlbw9PTEmDFjkJSUpH3e29sbKpUq2zJs2DDtPs2bN8/2/NChQ/V2j0REiYkyGerbF3BxAVq1ApYtk8mRvT3Qqxfw66+yh9rWrUCfPkyOiJSkaAnSpk2bEBAQgGXLlsHPzw8LFy6Ev78/rl69ChcXl2z7r1+/HhMmTMDKlSvRqFEj/PPPP+jfvz9UKhXmz58PADh9+jTS09O1x1y4cAH/+9//0LVrV51zDR48GDNnztQ+tra21tNdElFJFRsL7NolS4r27JFJUgYXF6BjR9me6J13OBI0UXGjaII0f/58DB48GAMGDAAALFu2DLt27cLKlSsxYcKEbPsfP34cjRs3xgcffABAlhb17NkTJ0+e1O5TJsuwr3PmzIGPjw+aNWums93a2hpubm6FfUtEVMJFR8tBG4OCgOBg3Uldvbwye541aiS7qxNR8aRYFVtKSgpCQ0PRsmXLzGBMTNCyZUuEhITkeEyjRo0QGhqqrYa7efMmdu/ejbZt2+Z6jbVr12LgwIFQZekDu27dOjg7O6N69eqYOHEiEp//akdElE+nT8tqsrJlgSFDgL17ZXJUrRowaRIQGiobYC9YALz1FpMjouJOsRKkmJgYpKenw9XVVWe7q6srrly5kuMxH3zwAWJiYtCkSRMIIZCWloahQ4fi888/z3H/bdu24cmTJ+jfv3+283h5ecHDwwPnz5/H+PHjcfXqVQQFBeUab3JyMpKfG642Li4uj3dKRMYqLU22K1q4EPjzz8ztvr5y4MZOneQo1kRkeAyqF9vhw4fxxRdf4LvvvoOfnx+uX7+OUaNGYdasWZgyZUq2/VesWIE2bdrAw8NDZ/uQIUO06zVq1IC7uztatGiBGzduwMfHJ8drBwYGYsaMGYV7Q0RkkGJjgR9/BBYtkvODAXIKjx49gNGjc59jjIgMh2IJkrOzM0xNTREVFaWzPSoqKte2QVOmTEGfPn3w4YcfApDJTUJCAoYMGYJJkybB5Lmx9cPDw3Hw4MEXlgpl8PPzAwBcv3491wRp4sSJCAgI0D6Oi4uDp6fnS89NRMbj+nXg22+BVauA+Hi5zdkZGDoU+OQTjlNEZEwUa4NkYWEBX19fBAcHa7dpNBoEBwejYcOGOR6TmJiokwQBgOl/FflCCJ3tq1atgouLC9q1a/fSWM6ePQsAcH/BXze1Wg17e3udhYiMnxDAoUNyoMbKlWWpUXy8nN7jxx/l7PKzZjE5IjI2ilaxBQQEoF+/fqhbty7q16+PhQsXIiEhQdurrW/fvihbtiwCAwMBAO3bt8f8+fNRp04dbRXblClT0L59e22iBMhEa9WqVejXrx/MzHRv8caNG1i/fj3atm0LJycnnD9/HmPGjEHTpk1Rs2bNort5IirWkpOBDRtk+6Jz5zK3t20LjBkDtGjB+c+IjJmiCVL37t3x8OFDTJ06FZGRkahduzb27t2rbbgdERGhU2I0efJkqFQqTJ48Gffu3UOZMmXQvn17zJ49W+e8Bw8eREREBAYOHJjtmhYWFjh48KA2GfP09ESXLl0wefJk/d4sERmEqCg5gON338ku+wBgbQ307w+MHAlUqaJoeERURFQia90U5UlcXBwcHBwQGxvL6jYiI3DuHPDNN8C6dUBKitz22mvAiBHAhx8CpUsrGx8RFY68fn4bVC82IqLCpNHIka4XLJDtjDL4+clqtM6dZe80Iip5mCARUYkTHw+sXi1LjK5fl9tMTYH335fd9Bs0UDI6IioOmCARUYkRHg4sXgwsXy7HMgIAR0c58vWwYUC5coqGR0TFCBMkIjJqQgAhIbIaLShIVqsBssv+qFFA376Ara2yMRJR8cMEiYiMUmoqsHmz7KZ/+nTm9pYtZTVamzaAiWIjwRFRcccEiYiMzsmTQLduchBHAFCrgd69ZYlRjRrKxkZEhoEJEhEZle3bgZ49gWfPAFdX2bZo6FCgTBmlIyMiQ8IEiYiMxpIlcjBHjUZWof3yC9sXEVHBsAaeiAyeRgN89hkwfLhcHzwY2LGDyRERFRxLkIjIoCUlyWlANm2Sj2fPBiZO5DxpRPRqmCARkcF6/Bjo1Ak4ehQwMwNWrgT69FE6KiIyBkyQiMgg3b4NtG0LXL4M2NvLMY5atFA6KiIyFkyQiMjghIUB7doBkZFyQtndu9l9n4gKFxtpE5FB2bMHaNpUJkc1a8pRspkcEVFhY4JERAZj+XKgfXsgIUGOiH30qCxBIiIqbEyQiKjYEwKYMkVOKpueLudP27ULcHBQOjIiMlZMkIioWEtJAfr1A/7v/+TjqVOB1asBCwtFwyIiI8dG2kRUbMXGAl26AMHBgKkp8P33wKBBSkdFRCUBEyQiKpbu3pXd+P/+W46IvXkz0Lq10lERUUnBBImIip3z52VydO8e4OYmu/HXqaN0VERUkrANEhEVKwcPAk2ayOSoWjXgxAkmR0RU9JggEVGxsWYN0KYN8PQp0KwZ8OefgJeX0lERUUnEBImIFCcEMGuWnHQ2LQ3o0QPYtw8oVUrpyIiopGKCRESKSk2V4xtNnSofjx8PrFsHqNXKxkVEJRsbaRORYp4+Bbp1A/buBUxMgEWLgE8+UToqIiImSESkkAcP5ISzZ84AVlbAxo1Ahw5KR0VEJDFBIqIid+mSbIwdEQGUKQPs3AnUr690VEREmdgGiYiK1JEjQOPGMjmqVAkICWFyRETFDxMkIioyGzcCrVoBT54AjRoBx48DPj5KR0VElB0TJCLSOyGAuXOBnj3l5LNdusgBIZ2dlY6MiChnTJCISK/S04Fhw2T3fQAYPRrYtEk2zCYiKq4UT5CWLFkCb29vWFpaws/PD6dOnXrh/gsXLkSVKlVgZWUFT09PjBkzBklJSdrnp0+fDpVKpbNUrVpV5xxJSUkYNmwYnJycYGtriy5duiAqKkov90dUkj15AnTuDCxdCqhUwIIFcjE1VToyIqIXUzRB2rRpEwICAjBt2jSEhYWhVq1a8Pf3R3R0dI77r1+/HhMmTMC0adNw+fJlrFixAps2bcLnn3+us98bb7yBBw8eaJdjx47pPD9mzBj89ttv2Lx5M44cOYL79++jc+fOertPopIoKAh4/XVgxw7A0hLYvFmWHhERGQJFu/nPnz8fgwcPxoABAwAAy5Ytw65du7By5UpMmDAh2/7Hjx9H48aN8cEHHwAAvL290bNnT5w8eVJnPzMzM7i5ueV4zdjYWKxYsQLr16/HO++8AwBYtWoVqlWrhhMnTqBBgwaFeYtEJc69e8Dw4cC2bfJx5cpyjjX+ahGRIVGsBCklJQWhoaFo2bJlZjAmJmjZsiVCQkJyPKZRo0YIDQ3VVsPdvHkTu3fvRtu2bXX2u3btGjw8PFChQgX06tULERER2udCQ0ORmpqqc92qVauiXLlyuV4XAJKTkxEXF6ezEFEmjQZYtkyWGm3bBpiZAZMnA+fOMTkiIsOjWAlSTEwM0tPT4erqqrPd1dUVV65cyfGYDz74ADExMWjSpAmEEEhLS8PQoUN1qtj8/PywevVqVKlSBQ8ePMCMGTPw1ltv4cKFC7Czs0NkZCQsLCzg6OiY7bqRkZG5xhsYGIgZM2YU/IaJjNiVK8DgwUBGbbafH7B8OVCjhrJxEREVlOKNtPPj8OHD+OKLL/Ddd98hLCwMQUFB2LVrF2bNmqXdp02bNujatStq1qwJf39/7N69G0+ePMEvv/zySteeOHEiYmNjtcudO3de9XaIDF5KCjBrFlCrlkyObGyAb74B/vyTyRERGTbFSpCcnZ1hamqarfdYVFRUru2HpkyZgj59+uDDDz8EANSoUQMJCQkYMmQIJk2aBBOT7Pmeo6MjKleujOvXrwMA3NzckJKSgidPnuiUIr3ougCgVquh5vTiRFohIbLU6OJF+bhtW9lbrVw5ZeMiIioMipUgWVhYwNfXF8HBwdptGo0GwcHBaNiwYY7HJCYmZkuCTP/rLyyEyPGY+Ph43LhxA+7u7gAAX19fmJub61z36tWriIiIyPW6RJTp6VNgxAg5XcjFi3IutfXr5XxqTI6IyFgo2ostICAA/fr1Q926dVG/fn0sXLgQCQkJ2l5tffv2RdmyZREYGAgAaN++PebPn486derAz88P169fx5QpU9C+fXttojR27Fi0b98eXl5euH//PqZNmwZTU1P07NkTAODg4IBBgwYhICAApUuXhr29PUaMGIGGDRuyBxvRS+zcCXz8MXD3rnzcvz/w1VeAk5OiYRERFTpFE6Tu3bvj4cOHmDp1KiIjI1G7dm3s3btX23A7IiJCp8Ro8uTJUKlUmDx5Mu7du4cyZcqgffv2mD17tnafu3fvomfPnnj06BHKlCmDJk2a4MSJEyhTpox2nwULFsDExARdunRBcnIy/P398d133xXdjRMZmKgoYNQoOQI2AFSoAHz/PfBcZ1AiIqOiErnVTdELxcXFwcHBAbGxsbC3t1c6HCK9EAJYvRr49FPg338BExO5Pn06YG2tdHRERPmX189vRUuQiKj4un4d+Ogj4Pff5eM6dYAffwTefFPZuIiIioJBdfMnIv1LTQW+/FJ20//9dzmp7Ny5wKlTTI6IqORgCRIRaYWGAh9+CJw9Kx+3aCHbGvn4KBoWEVGRYwkSESEhQbYtql9fJkelSgGrVgEHDjA5IqKSiSVIRCXc/v2yrdHt2/Jxz57AwoWAi4uSURERKYslSEQlVEwM0Lcv4O8vkyNPT2DXLjnoI5MjIirp8p0geXt7Y+bMmYiIiNBHPESkZ0IA69YB1aoBP/8MqFTAyJFyVOy2bZWOjoioeMh3gjR69GgEBQWhQoUK+N///oeNGzciOTlZH7ERUSG7fVsmQb17yxKk6tWB48flBLN2dkpHR0RUfBQoQTp79ixOnTqFatWqYcSIEXB3d8fw4cMRFhamjxiJ6BU9fQrMnw+88Qawdy9gYQHMmiV7rXGGHSKi7F55JO3U1FR89913GD9+PFJTU1GjRg2MHDkSAwYMgEqlKqw4ix2OpE3F3aNHwI4dwNatsiF2RkHvW28BP/wAVK2qbHxERErQ+0jaqamp2Lp1K1atWoUDBw6gQYMGGDRoEO7evYvPP/8cBw8exPr16wt6eiLjIgTw5AkQHg5ERgLOzoCXl/y/EL9I3LsHbNsGBAUBR44A6emZz1WsCIwdCwweLKcMISNz9aocn4Et7IkKRb4TpLCwMKxatQobNmyAiYkJ+vbtiwULFqDqc19HO3XqhHr16hVqoETFWno6cP8+EBEhk6CM/59fj4/PfpyVFVCunEyWsv7v5QWULQuYm7/w0teuyVKioCDg5End52rVAjp3lssbbxRqLkbFyYYNQK9egL098MsvQKtWSkdEZPDynSDVq1cP//vf/7B06VJ07NgR5jn88S5fvjx69OhRKAESFQuJiTLRySnxiYgA7t4F0tJefh4XF8DNTbaQfvAAePZMfvO/ejXn/U1MAA8PnaRJeJbDLY0X9lwsh5+OeOHUZd3W1Y0ayYSoUyegQoVCuHcq3nbvluM1CAHExspW+AsXAsOHKx0ZkUHLdxuk8PBweHl56Sseg8E2SEZECNlgJ6fEJ2NbTMzLz2NmJgcTer4E6PlSoXLlZIlRhuRkmVjllnRFRGQ2HHqBxyiFf23LwaSCF8r4loPt61mu7eLCOjVj9ccfsrQoKQn44ANZ2rhmjXzuk09kovSSEkiikiavn9/5TpBOnz4NjUYDPz8/ne0nT56Eqakp6tatW7CIDQwTJAN35QowcaL8PyJClhC9jJ1dzolPxjY3N8DUtNBCTE3WIGR7NEI2ReDG7+GwexIBL4TLRRUBH7Nw2KX++/ITqdUycatQQTZC+t//Ci1GUtDZs0CzZkBcHNCunaxnNTMD5s0DJkyQiX/LlrLKrVQppaMlKjb0liDVr18fn332Gd5//32d7UFBQfjyyy9xMmsjCCPFBMmAJSYCdeoA//yju93dPefEJ2PdwUHvjXiePZM9zoKCgN9+A/59Lv+xtwfefVdWn7VuDdjYQPbff1G7p/v3AY0m8yRmZnKo7K5d9XofpGfXrgFNmgDR0UDTpnLshudLJ7dvl22SEhKAypWBnTuBSpWUi5eoGNFbgmRra4vz58+jQpbGDbdu3ULNmjXx9OnTgkVsYJggGbBhw4DvvpNte9asAby9ZQmLWq1IOLGxcoqPrVtlc5LnC7PKlAE6dpRJ0dtvFyDE1FTZtS08HFi6FNi0SVa3rVwJ9OtXmLdBReXuXZkchYfLRP/QIZm8Z3XuHNC+PXDnjixB+vVX4J13ij5eomJGb9381Wo1oqKisiVIDx48gJkZ576lYm7PHpkcAcDq1bIKQgHR0XKMoqAg4OBBmcdkKFcus5F148avWGtnbi4TQG9v+aFqawusWAH07y8zsY8/frUboaIVEyPbHIWHy5KhvXtzTo4A2YXx1CmZYZ88KSfdW7IEGDKkSEMmMlT5LkHq2bMnHjx4gO3bt8Phv1/MJ0+eoGPHjnBxccEvv/yil0CLG5YgGaBHj+TcGpGRcvKxb74p8stv3gxs3Cjb1j5f81WtmkyIOncG3nxTjzV5Gg0wZgzw7bfy8VdfAZ9+qqeLUaF6+lSWAP31F/Daa8Cff8ps+mWSkoBBg2TVKgCMGiV/7vxCSyWU3qrY7t27h6ZNm+LRo0eoU6cOAODs2bNwdXXFgQMH4Onp+WqRGwgmSAZGCNnuZssWmY2Ehuq22dCTpCTZ/GPtWll99nxJka9vZklRtWp6DyWTEMCkSUBgoHw8YwYwZQoHSSrOkpJk9/1Dh+Tgon/8kb+h0IUAvvgCmDxZPm7TRo6dlFvpE5ER01uCBAAJCQlYt24dzp07BysrK9SsWRM9e/bMcUwkY8UEycD89JNsc2NmJqsb3nxTb5fSaICjR2VS9Ouvso1Rhtq1ZdvZrl1lu29FzZ6d+YH52WfAnDlMkoqjtDTg/fdlw2s7O5kk+foW7Fy//irHTHr2DHj9ddkTgINlUQmj1wSJmCAZlNu3gZo1ZRXF7NnA55/r5TIXLsikaP162S42g6enTIp69ZI1fMXKwoWyyg2Qjde//ZZjJhUnGg0wcKDsTKBWyzZHzZu/2jlDQ4EOHWQPRycn2RCuadNCCZfIEOg9Qbp06RIiIiKQkpKis71Dhw4FOZ3BYYJkINLTZfevP/6QQ0wfPVqoYxXduydrKtaulZ2GMjg4yFKi3r3l5LDFOuf44Qdg6FBZDdO/P/Djj4X6GlEBCQEEBMgk1tRUJjKF9ff13j3gvfdksmRuDnz/PTBgQOGcm6iY01svtps3b6JTp074+++/oVKpkJFfqf4rmk9/fnZMIqXNny+TI1tb4OefC+WDPy5OflatXQv8/rv8HAPk50y7djIpatcOsLR85UsVjSFDAGtrWQW5erWsfvn5Z47ArLTZs2VyBACrVhVecgTIOf6OHpU/819/laVUly/LdmlMjokAAPn+Xjtq1CiUL18e0dHRsLa2xsWLF3H06FHUrVsXhw8f1kOIRAV07pxsjAzID5pXaGuRmiobW/foAbi6yi/bwcEyOWrSBFi2THaO27oV6NLFgJKjDL17yxGXzc3lWEnvvy8bBpMyliyRDecB2duyT5/Cv4a1tfxZT50qH8+bJ3sNlJCx7IheSuSTk5OTOHfunBBCCHt7e3HlyhUhhBDBwcGidu3a+T2dwYqNjRUARGxsrNKhUE6ePROienUhACHee08IjSbfp9BohAgJEWLYMCGcnOSpMpYqVYT4v/8T4ubNwg9dUbt2CaFWy5v83/+ESEhQOqKSZ+3azDfa9OlFc8316zN/7jVrCnH7dtFcl0gBef38zncJUnp6Ouzs5Ozhzs7OuH//PgDAy8sLV3ObkZyoqE2eLFtNu7jINjb56J117RowfbqcmaFhQ/ll/tEjWXI0erQchubyZVk4Vb683u5AGW3byvEIbGyAAwfknCZxcUpHVXLs3Jk5wvmIEZmlO/rWsydw5Ih8k58/D9SvD4SEFM21iYqpfLdBql69Os6dO4fy5cvDz88Pc+fOhYWFBX744Ydso2sTKeLQIdn2CJCjRru4vPSQ6GhZ27B2rRx8OIO1tax16N0baNGihIyt9847ckK4Nm1k+63//U+OQF66tNKRGbejR2XL/vR0+YZbuLBoh13w8wNOn5bTk5w7J3vLrVwpu18SlUD57sW2b98+JCQkoHPnzrh+/Treffdd/PPPP3BycsKmTZvwTgmZ64e92IqpJ09kl/47d2Tj4++/z3XXxEQ5tMzatcC+ffJzCZA9zlq1kp9R770n23eXSKGh8oV4/Fi+pgcO5CnZpAI4c0YmJHFxMkHZskW5RvLx8fLNv327fDxpEjBzZjHvilkEEhMzJ4HWaOQo5uXKybGpyKAU6ThIjx8/RqlSpbQ92UoCJkjFVO/ewLp1QMWK8kMnh+zm3j3g66+B5cvlZ0GGunXl4d27A25uRRhzcXbhgpyvLioKqFJFtkwvW1bpqIzLP//Ilv4PHwLNmsnSuiIY5f2FNBqZGM2ZIx937iwHW7WxUTYufRFCznOXkQCFh2euZ/wfE5PzsaVKyVFfy5XT/T9j3dWVA7AWM3pJkFJTU2FlZYWzZ8+ierEb8a5oMUEqhjZtkt3MTEzkPFUNGug8ffMm8OWXsid7xvBd3t4yKerVK38zN5Qo167J+sU7d2Sjq+BgI2x8pZA7d2RyFBEhR3c/dAgoTn9PfvoJGDxY/sK8+aYsVXrtNaWjyr/UVPnNKKfEJ2P92bOXn8fOTiY+pqbymH//ffkxarUcLfb5pOn5dU9PwMLi1e+R8kwv4yCZm5ujXLlyhTrW0ZIlSzBv3jxERkaiVq1aWLRoEerXr5/r/gsXLsTSpUsREREBZ2dnvP/++wgMDITlf/2qAwMDERQUhCtXrsDKygqNGjXCl19+iSpVqmjP0bx5cxw5ckTnvB999BGWLVtWaPdFRezevcyZ6SdN0kmOLl6UX4Q3bMisRnvrLTmgtr8/v9y9VKVKsi3SO+/ILLNpU5kkVa6sdGSG7eFDWYUZESFL5/buLV7JESCnJfHxkRMGhoXJxtvbtwP16ikdma74+NxLfsLD5ajhz88OnRt39+wJzPPrjo66+z99mnOpU8b6/ftAcjJw/bpccqJSySLrrNd8/tqcM08Z+e0e9+OPP4q2bduKR48eFaBzna6NGzcKCwsLsXLlSnHx4kUxePBg4ejoKKKionLcf926dUKtVot169aJW7duiX379gl3d3cxZswY7T7+/v5i1apV4sKFC+Ls2bOibdu2oly5ciI+Pl67T7NmzcTgwYPFgwcPtEt+u+uzm38xkp4uu6QDQtStK0RKihBCiNOnhejUSbd7fuvWQhw9qnC8hurePSGqVZMvpKurEOfPKx2R4YqNFcLXV76Wnp5CREQoHdGL3bolxBtvyHgtLYXYtKnorq3RCPHggRAnTwrxyy9CfPWVECNGCNGhgxC1awtRqpTuL3lui4WFEBUrCvHOO0IMGCCHUFi5UojgYCGuXRMiKanwY09Jka/d4cNCrFkjxKxZQnz4ofx7VaWKfC3zEru9vRA1agjRvz+HYCgEef38zncbpDp16uD69etITU2Fl5cXbLLUSYeFheX5XH5+fqhXrx4WL14MANBoNPD09MSIESMwYcKEbPsPHz4cly9fRnBwsHbbp59+ipMnT+LYsWM5XuPhw4dwcXHBkSNH0PS/+YaaN2+O2rVrY2HGKLUFwCq2YmTRImDkSMDKCiLsDP6IroLZs2VHLEB+QevcGZg4seBzfNJ/Hj6UvdrOnZO92vbtk423KO+ePZM9BI8cAcqUkaVzz5VwF1txcXI4gN275eMZM+Rglq9aBJuSAty9m3spTESELIV5GUfHl7cFKm4NzTPaPuV27+HhcoyR51lZyeLvsWMNcETa4kFvU4107NjxVeLSSklJQWhoKCZOnKjdZmJigpYtWyIkl/E3GjVqhLVr1+LUqVOoX78+bt68id27d6PPC0aZjf1vKvXSWboor1u3DmvXroWbmxvat2+PKVOmwNraOtfzJCcnI/m5X9I4jg1TPFy6JGeiB3Cx/zx89GEV/PmnfMrUFPjgA2DCBDlxORWCMmVkO5k2bYCTJ2XbpN27gcaNlY7MMKSmyl4AR47I9ix79xpGcgTI6r8dO+Tv2/z5wLRpckCwlStf3Kg8NvbF1VAPHmTO15MblQrw8HhxAmSIX1RVKvk7VaZM7l80EhLk63Tzphzt/MgRmZiuWSMnl27TpmhjLkmKpDwrB/fu3RMAxPHjx3W2jxs3TtSvXz/X47755hthbm4uzMzMBAAxdOjQXPdNT08X7dq1E40bN9bZ/v3334u9e/eK8+fPi7Vr14qyZcuKTp06vTDeadOmCQDZFlaxKSg5WWjefFMIQPxp5y8AjbYkfehQIW7cUDpAIxYXJ0TTprL439paiIMHlY6o+EtPF6JPn8xqqsOHlY6o4JYvF8LMTN5LvXpCHDokxMaNQnz5pRx6/t13ZZWQg0PeqpAsLYWoXFlWPQ0aJMTMmbJK6tAhOVx9crLCN1xMaDRCrFsnhLt75mv33nuyGo/yLK9VbAaVIB06dEi4urqK5cuXi/Pnz4ugoCDh6ekpZs6cmeP+Q4cOFV5eXuLOnTsvjCU4OFgAENevX891n6SkJBEbG6td7ty5wwRJQSkpQpxrP0kIQMSgtHDHPWFjI8Snn8qmMlQEEhKEaNVK/pFWq4XYuVPpiIovjUaIkSPla2VqKsRvvykd0as7dEiI0qXzlgA5OQlRp44QHTsKMWqUEF9/LcSvvwpx6pQQUVEFmgqoRIuNlX/sMpJUS0shZsyQUyzRS+ktQVKpVMLExCTXJa+Sk5OFqamp2Lp1q872vn37ig4dOuR4TJMmTcTYsWN1tv3888/CyspKpKen62wfNmyYeO2118TNPEyWFR8fLwCIvXv35jl+NtJWxrNnQnz3nRCd3f4UaTARAhB9rTeLKVOEiIlROroSKClJfoMF5B/rX35ROqLiacaMzGRh7Vqloyk8164J0ayZbGjepIkQvXoJ8fnnQixbJsSePUJcvCjE06dKR2m8Ll4U4u23M99bFSoYR/KtZ3n9/M53G6StW7fqPE5NTcWZM2ewZs0azJgxI8/nsbCwgK+vL4KDg7XtmjQaDYKDgzF8+PAcj0lMTIRJlkZ2pqamAADxXx22EAIjRozA1q1bcfjwYZTPw3gtZ8+eBQC4u7vnOX4qWk+fykGxv/4aiI98irPoA1No8PebfbHo0PsG2fzAKKjVwObNsjv4xo1yHKpnz+RjkhYtku11MtaNaeqOihWBw4eVjqLkev11OeTGL78AAQGynVL79sC77wLffANw+q9XU1gZ2bp163It+cnNxo0bhVqtFqtXrxaXLl0SQ4YMEY6OjiIyMlIIIUSfPn3EhAkTtPtPmzZN2NnZiQ0bNoibN2+K/fv3Cx8fH9GtWzftPh9//LFwcHAQhw8f1unGn5iYKIQQ4vr162LmzJnir7/+Erdu3RLbt28XFSpUEE2bNs1X7CxBKhqPHsneuM/35F1v86EQgEj3LCfEkydKh0hCCJGWJsTAgZk/pKVLlY6oePj558zXJJemAESF4ulTIT77LLPaTa0WYupUIf777KNMRd4G6caNG8LGxibfxy1atEiUK1dOWFhYiPr164sTJ05on2vWrJno16+f9nFqaqqYPn268PHxEZaWlsLT01N88skn4t9//9XugxwaUgMQq1atEkIIERERIZo2bSpKly4t1Gq1qFixohg3bhzHQSpmHjwQYtw4IWxtMz9fKlUSYv+I7fKBSmXYjVyNUXq6HJ8m4wf29ddKR6SsHTtkeyNAtrthOxsqCpcvC9GyZebvobe3ENu28f33nCJNkBITE8WoUaNE5cqVC+N0BoEJkn7cvi3EJ5/ILz8Zv981a8oOMmn3IoUoU0ZuHDdO6VApJxqNEBMm6JaalMQ/zIcOZb6J+/aVySNRUdFohNi8WYjXXsv8XWzTRrYZI/0lSI6OjqJUqVLaxdHRUZiamgo7Ozuxffv2AgdsaJggFa4rV+QgsRmlw4AQDRrI9oYajZD/vPtuZsakj1FvqXBoNHLE4Iwf5PjxJStJ+usvIezs5L136CBEaqrSEVFJFR8vv7CYm2eOJj55suyBWoLpbSTt1atXQ/XcyKkmJiYoU6YM/Pz8UKpUqUJoFWUYjG4k7XnzgMWLZSPbgAA56mwRuHcPGDMG+PXXzLHiWrSQA8W+/fZzg/QuXw4MGSIndfzrL6BGjSKJj17BggXyvQTIn93ixYC5ubIx6dvJk7KBbEyMfAPv3s3Rjkl5V6/K2QYyphfw8pK/nx07lsjJKPP8+V0k6ZoRMqoSpG+/zT5o27Bhep/zJzlZiP/GedR+2Q4JyWHHa9eEsLGRO331lV5jokK2bJlsLwbIObAKYQ7HYmv9+sxqtbp15Vg1RMWFRiPEli1ClCuX+UfX31+Iq1eVjqzI5fXzO98T06xatQqbN2/Otn3z5s1Ys2ZNfk9HSlu7Vn6zAIAPPwT8/ICkJGDJEtmFd8AA+e1DDyZPlhOEly4NnDkjJwlv0CDLTmlpQJ8+crj95s1lcRMZjo8+ArZtA2xsgN9/lz/gf/5ROqrCpdHIbvwffCDnDGvfXt6rMZQsk/HImJTy8mVg0iRZGr9vnyyN//xz+TeWdOU386pUqZL4/fffs20/fPgwG2kbmpx62Wg0cnbrFi0yv2WoVEJ07SrEmTOFdukDBzJPn2WsUF0ZbVns7YUIDy+061MRO3cu85uro6PxTE2SkCBEt26Zb+Zx4+SQB0TF3T//yIbbGe9dT0/ZsLsEtBfUWyNttVotbuUw78utW7eEpaVlfk9nsAw+QcpLL5uQEFnv9Xz1W9u2Qhw79kqXfvgwcyqhjz56wY6nT2e22v7551e6JhUDkZFCNGyYOd3GsmVKR/Rq7t2TVWmAbAS7cqXSERHlj0YjhwDw9s78G9+ypRwqwIjprYrNxcUF58+fz7b93LlzcHJyeuUSLSoCoaFAhw6yOqBDB2DFCsAkh7dCgway3uv8eaBnT7nP7t1Akyayumv//syW1XkkhKzJe/AAqFpVTgqeo8REoHdvWcXWrZtxjT5cUrm6yqqnXr2A9HRg6FBg1Cj5MzY0oaFAvXqyw4CTkxzNeMAApaMiyh+VCnjvPeDSJWDqVDky/sGDQM2awPjxQHy80hEqK7+Z12effSa8vLzE77//LtLS0kRaWpoIDg4WXl5e4tNPPy1wRmdoDLYE6fJlIZyd5TeF5s3zN7nhtWtCfPhhZpfRjMaoQUF5Hudl2bLML9xhYS/YcfhwuaO7u3E37C2JNBoh/u//dBuKGtKI6Js3C2FlJWN//XUhbtxQOiKiwnH9euZwKoAQZcsKsWmT0VW76a2KLTk5WXTr1k2oVCphbm4uzM3NhampqRgwYIBITk4ucMCGxiATpPBwWc8MCOHrW/BeNnfuyDZLGR8SGR8UP//8wjFfLl3KPOSFndH27s087759BYuRir9ff818Q1SrJv84F2dZx3dq3dqwEjuivNqxQ4jy5XW/xBjR2El6H0n7n3/+Eb/88ov47bffxG09dwcvjgwuQYqKEqJyZflmr1pViOjoVz9ndLScudvePvMXqXx5WUyUZSDHpCQhatWSu/zvfy8ocIqJyWygNGLEq8dIxVtoqPyWCgjh5CTEkSNKR5SzZ8+E+OCDzPf56NEcAJKMW2KinAjT0tLo5hIs8rnYShqDSpCePBGiTh35Ji9XToiIiMI//+zZmVV3gBAeHnIurvh4IYQQAQFys7OzEPfv53IejUaI99/PTOKM6BsLvUDWxs4rVigdka7ISDmsOyA7DRh643Ki/NiwQb73bWzkJJlGQG+NtLt06YIvv/wy2/a5c+eia9eur9QeivTg2TPZEPvMGaBMGeDAAcDTs3Cv4eAgx9EIDwcWLgTKlgXu3wc+/RTw8sL1frOwcv6/AGR7cHf3XM6zdq0cUtvMTK5bWxdunFQ8eXgAR47IxvipqcCgQcC4cbIht9LOnZONsU+cAEqVkuPGfPSR0lERFZ3u3eX4eAkJsiF3SZLfzMvZ2VmcP38+2/bz588LFxeX/J7OYBlECVJKSmaDO3v7l7SKLkRJSUIsXy6Ej4+2RCkWdmJvnfHy23hObt/OrKr7v/8rmjipeNFohJg2LbMU8t13hYiLUy6e7dszR3CvXLlEjjhMJISQQ7sAQpiYCPH330pH88r0VoIUHx8PCwuLbNvNzc0RFxdXCCkbFQqNRnY73rlTzgX1229AnTpFc221GvjwQ4jLVzC3zgacRw3Y4yn8z3wJeHsDI0YAERGZ+6enA/36AXFxQMOGsnsplTwqFTB9OrBhg3wP7dwJNG4sSyaLkhDA3LlynqqEBDk54IkTQOXKRRsHUXHRuDHQpYv8XBk3Tuloiky+E6QaNWpg06ZN2bZv3LgRr7/+eqEERa9ICGD0aGDdOlld9euvQNOmRR7G0uVmGH+mB+qbn8Wtb3ZkTmOyeDHg4wMMHCinMVmwQFax2NgAP/8sY6aSq0cP+X5wcwP+/huoXx84frxorp2cLN+X48fL36OPPwb27JHVa0Ql2Zdfysmm9+7NnPTW2OW3aGrHjh3CzMxM9O3bV6xevVqsXr1a9OnTR5iamoqtL5wzwrgU6yq2jGoKlUqIdesUCeHChczODwsW/LcxYxqTd97RncYkY7Ts5csViZWKqYgIIWrXlu8NCwshfvpJv9eLjhaiSZPMqoRFi/R7PSJDM2aM/P2oUcOgp9TRay+2nTt3ikaNGglra2vh5OQk3nnnHXHkyBHxtxHUTeZVsU2QFi7MTD4WL1YkhGfP5O9PxvAZOXbpDwkRon37zFg7dDC6wcioEMTHC9GpU+b7ZOLEPA9Kmi8XLmSO+2JvL8fiIiJdjx4JUaqUwX+hLbJu/rGxsWLZsmWiXr16wsTE5FVPZzCKZYK0Zk3mB8msWYqFMWqUDKFMmTz0Cj13Tohvvy34oJVk/NLT5XhbGe/tTp20w0cUit27hbCzk+euUEGOaEpEOZs/X/6uuLkJ8fSp0tEUiN4aaWc4evQo+vXrBw8PD3z99dd45513cOLEicKp96P8275dtp0AgDFjgEmTFAlj717gm2/k+qpVshnJC9WsKRtt29vrPTYyUCYmwOzZwE8/ARYWwNatwFtvAXfvvtp5hZDDUrz7LvD0KdCsGXDqFFCtWqGETWSUhg2TbUgjI4F585SORq/ylSBFRkZizpw5qFSpErp27Qp7e3skJydj27ZtmDNnDurVq6evOOlFDh2SY1Vk9Ab76ivZI6iIRUXJywPA8OFAu3ZFHgIZsz595GS3ZcrIcb3q1ZMJTUGkpMjxjMaMkT1zBg2SDU854TbRi1lYAHPmyPV584B795SNR4/ynCC1b98eVapUwfnz57Fw4ULcv38fixYt0mdslBd//SUHgkxOlrMy//ij/MZdxISQBVjR0UD16rKXNFGha9xYJkXVq8tvsM2aATn0qn2hR48Af39g+XL5ReLrr+V6DsOXEFEOunSRv4vPngFTpigdjd7k+ZN0z549GDRoEGbMmIF27drB1NRUn3FRXly+DLRuDcTHA2+/DWzcqFgX+cWLgd275fA169cDVlaKhEElgbe37Pb/7rty2IgePeT4SUK8/NgrV4AGDYDDhwFbW2DHDiAgQJESVyKDlfHFAgBWrwbOnlUyGr3Jc4J07NgxPH36FL6+vvDz88PixYsRExOjz9joRcLDgVat5LfhevVkGyRLS0VC+fvvzLHD5s0DatRQJAwqSezsgG3bgLFj5eMZM2Si9OxZ7sccOCCTo+vXAS+vzCSLiPLPz0827RBC/h7m5QuKgclzgtSgQQMsX74cDx48wEcffYSNGzfCw8MDGo0GBw4cwNOnT/UZJz0vOhr43/9kI9Vq1WTRjZ2dIqE8ewZ88IGs4WvbVrY9IioSpqYyI1+xQg5g98svssrtwYPs+373HdCmDRAbCzRqJKvpmMkTvZrAQFk1HRwsB1Q1MvlurGJjY4OBAwfi2LFj+Pvvv/Hpp59izpw5cHFxQYcOHfQRIz0vNla2n7h2TX4L3r8fcHZWLJzx44ELFwAXF9lrjTUVVOQGDpSlQ6VLA6dPyxLVM2fkc2lpMmsfNkx2YujbVzb0dnFRNmYiY1C+PDBqlFwfO1b+vhmRV2rNW6VKFcydOxd3797Fhg0bCismyk1iItC+vazvdXGRHwqvvaZYOLt2ARnt9Nes4WcOKSiji37VqrJXTZMmcliAtm2BJUvkPoGBsr2EWq1oqERG5fPPZe/Py5dlJyEjohLCCCsOi0BcXBwcHBwQGxsL+6IYwyc1VU6euXu3HDPoyBGgdm39XzcXkZFyCKOHD+UXiIULFQuFKFNsrGwXsW9f5jZra2DtWqBTJ+XiIjJmixYBI0fKITiuXy/249rl9fO76PuDU/5pNED//jI5srKSRTcKJkcaDTBggEyOatbMHBKDSHEODsDOnfKPNSBLWI8dY3JEpE9DhwKVK8sPhS+/VDqaQsMEqbgTQv6xX79eduH/9VdZfaCgRYvkiNmWljIshTrPEeXMzEwO5372LHDxIlCnjtIRERk3c/PMwe/mzwfu3FE2nkLCBKm4mzZNtqFQqTLbVCjo3Dngs8/k+tdfA2+8oWg4RLmrVavYF/UTGY0OHYCmTeXYZJ9/rnQ0hYIJUnG2YAEwa5ZcX7IE6NlT0XAyuvSnpMi24h9/rGg4RERUXDw/eOTatUBoqLLxFALFE6QlS5bA29sblpaW8PPzw6mXzK20cOFCVKlSBVZWVvD09MSYMWOQlJSUr3MmJSVh2LBhcHJygq2tLbp06YKoqKhCv7dXsnq1HOEXkBN1FoNsZOxY4NIlOQHtihXs0k9ERM+pWxfo3Vuuf/qp4Q8eKRS0ceNGYWFhIVauXCkuXrwoBg8eLBwdHUVUVFSO+69bt06o1Wqxbt06cevWLbFv3z7h7u4uxowZk69zDh06VHh6eorg4GDx119/iQYNGohGjRrlK/bY2FgBQMTGxhbs5l9k61YhTEyEAIQICBBCoyn8a+TTjh0yHECIffuUjoaIiIql8HAhLC3lh8W2bUpHk6O8fn4rmiDVr19fDBs2TPs4PT1deHh4iMDAwBz3HzZsmHjnnXd0tgUEBIjGjRvn+ZxPnjwR5ubmYvPmzdp9Ll++LACIkJCQPMeutwTp4EEhLCzkm2vAgGKRHN2/L4SzswzpuVyUiIgou4kT5QdG5cpCpKQoHU02ef38VqyKLSUlBaGhoWjZsqV2m4mJCVq2bImQkJAcj2nUqBFCQ0O1VWY3b97E7t270fa/hst5OWdoaChSU1N19qlatSrKlSuX63UBIDk5GXFxcTpLoUtJkaMCp6TIbsk//KB4PZZGA/TrB8TEyDavgYGKhkNERMXdhAly5OB//gG+/17paApMsQQpJiYG6enpcHV11dnu6uqKyMjIHI/54IMPMHPmTDRp0gTm5ubw8fFB8+bN8fl/Lebzcs7IyEhYWFjA0dExz9cFgMDAQDg4OGgXT0/P/N7yy1lYyDGOevXK7NavsIUL5YDdVlbAhg0chJiIiF7C3l5OIA0A06cDT54oGU2BKd5IOz8OHz6ML774At999x3CwsIQFBSEXbt2YVZGTy89mjhxImJjY7XLHX2N81C9uuwBUAwGFzpzRn4RAOTQFtWqKRsPEREZiA8/lB8ajx4BX3yhdDQFolgRhbOzM0xNTbP1HouKioKbm1uOx0yZMgV9+vTBhx9+CACoUaMGEhISMGTIEEyaNClP53Rzc0NKSgqePHmiU4r0ousCgFqthroEFZ8kJsou/ampwHvvAR99pHRERERkMMzMgHnzgHfflQO3fvIJ4O2tdFT5olgJkoWFBXx9fREcHKzdptFoEBwcjIYNG+Z4TGJiIkxMdEM2NTUFAAgh8nROX19fmJub6+xz9epVRERE5HrdkiggALhyBXB3l/MPsks/ERHlS9u2QIsWsl3txIlKR5N/RdNmPGcbN24UarVarF69Wly6dEkMGTJEODo6isjISCGEEH369BETJkzQ7j9t2jRhZ2cnNmzYIG7evCn2798vfHx8RLdu3fJ8TiFkN/9y5cqJ33//Xfz111+iYcOGomHDhvmKXa/d/BW2dWtml/4DB5SOhoiIDNaZM0KoVPID5cQJpaMRQuT981vRVsDdu3fHw4cPMXXqVERGRqJ27drYu3evtpF1RESETonR5MmToVKpMHnyZNy7dw9lypRB+/btMXv27DyfEwAWLFgAExMTdOnSBcnJyfD398d3331XdDdejN27BwwaJNfHjgWe6+xHRESUP7Vry67QGYMfHztmMFUSKiEMfahLZcTFxcHBwQGxsbGwN5L5njQaoFUrIDhYzu954oTsWEdERFRg9+4BlSrJ+ap+/RXo0kXRcPL6+W1QvdhIv77+WiZH1taySz+TIyIiemVly8oqCQAYP162STIATJAIABAWBkyaJNcXLgSqVFE0HCIiMiaffSYn8rxxQ06+bgCYIBEAYNYs2aW/Uyc5fAUREVGhsbWVHzSA/P/xY2XjyQMmSIT0dODQIbn++ecG036OiIgMyYABcjDkf/8F/u//lI7mpZggEcLCgNhYwMFBNs4mIiIqdKamwFdfyfXFi4Hr15WN5yWYIBF+/13+37y5fP8SERHphb+/XFJTi/3gkUyQSJsgvfOOsnEQEVEJ8NVXgImJ7PL/559KR5MrJkglXEoK8Mcfcp0JEhER6V316sDAgXL900/lxA3FEBOkEu7kSTl2V5kywBtvKB0NERGVCLNmATY28kNo0yalo8kRE6QS7vnqNfZeIyKiIuHmJgeNBIAJE4CkJGXjyQETpBKO7Y+IiEgRAQGAhwcQHg4sWqR0NNkwQSrBEhOBkBC5zgSJiIiKlI0NkDHZ/OzZQEyMsvFkwQSpBPvzT9nTslw5wMdH6WiIiKjE6dsXqF1bDsY3c6bS0ehgglSCBQfL/9n+iIiIFGFiImdKB4ClS4F//lE2nucwQSrB2P6IiIgU9847QLt2QFqanNS2mGCCVEI9eQKEhsr1t99WNBQiIirp5s2TUzls3w4cOaJ0NACYIJVYR48CGg1QuTLw2mtKR0NERCVatWrAkCFy/dNP5QeUwpgglVCsXiMiomJl+nTAzk5Wb6xfr3Q0TJBKKiZIRERUrLi4ZE5g+/nncpoHBTFBKoGio4G//5brzZsrGgoREVGm0aMBT0/gzh1g4UJFQ2GCVAIdPiz/r1lTzsFGRERULFhZAV98IdcDA+U3eoUwQSqBMqrXWrRQNg4iIqJsPvgAqFsXePoUCApSLAwzxa5Minl+gEgiIqJixcQE+O47OR9Ws2aKhcEEqYSJiACuX5fDTTRtqnQ0REREOahXT+kIWMVW0hw6JP+vWxewt1c2FiIiouKKCVIJw+79REREL8cEqQQRggkSERFRXjBBKkGuXwfu3gUsLIBGjZSOhoiIqPhiglSCZJQeNWwIWFsrGwsREVFxxgSpBOH4R0RERHnDBKmE0Ggye7Cx/REREdGLFYsEacmSJfD29oalpSX8/Pxw6tSpXPdt3rw5VCpVtqVdu3bafXJ6XqVSYd68edp9vL29sz0/Z84cvd6nki5cAB4+BGxsisXwEkRERMWa4gNFbtq0CQEBAVi2bBn8/PywcOFC+Pv74+rVq3Bxccm2f1BQEFJSUrSPHz16hFq1aqFr167abQ8ePNA5Zs+ePRg0aBC6dOmis33mzJkYPHiw9rGdnV1h3Vaxk1G99tZbspE2ERER5U7xBGn+/PkYPHgwBgwYAABYtmwZdu3ahZUrV2LChAnZ9i9durTO440bN8La2lonQXJzc9PZZ/v27Xj77bdRoUIFne12dnbZ9jVW7N5PRESUd4pWsaWkpCA0NBQtW7bUbjMxMUHLli0REhKSp3OsWLECPXr0gI2NTY7PR0VFYdeuXRg0aFC25+bMmQMnJyfUqVMH8+bNQ1paWq7XSU5ORlxcnM5iKNLSgCNH5DoTJCIiopdTtAQpJiYG6enpcHV11dnu6uqKK1euvPT4U6dO4cKFC1ixYkWu+6xZswZ2dnbo3LmzzvaRI0fizTffROnSpXH8+HFMnDgRDx48wPz583M8T2BgIGbMmJGHuyp+wsKAuDjA0RGoXVvpaIiIiIo/xavYXsWKFStQo0YN1K9fP9d9Vq5ciV69esHS0lJne0BAgHa9Zs2asLCwwEcffYTAwECo1eps55k4caLOMXFxcfD09CyEu9C/jOq15s3lJLVERET0YopWsTk7O8PU1BRRUVE626Oiol7aNighIQEbN27Mseoswx9//IGrV6/iww8/fGksfn5+SEtLw+3bt3N8Xq1Ww97eXmcxFGx/RERElD+KJkgWFhbw9fVFcHCwdptGo0FwcDAaNmz4wmM3b96M5ORk9O7dO9d9VqxYAV9fX9SqVeulsZw9exYmJiY59pwzZMnJwLFjcp0DRBIREeWN4lVsAQEB6NevH+rWrYv69etj4cKFSEhI0PZq69u3L8qWLYvAwECd41asWIGOHTvCyckpx/PGxcVh8+bN+Prrr7M9FxISgpMnT+Ltt9+GnZ0dQkJCMGbMGPTu3RulSpUq/JtU0MmTwLNngKsrUK2a0tEQEREZBsUTpO7du+Phw4eYOnUqIiMjUbt2bezdu1fbcDsiIgImJroFXVevXsWxY8ewf//+XM+7ceNGCCHQs2fPbM+p1Wps3LgR06dPR3JyMsqXL48xY8botDEyFs9Xr6lUysZCRERkKFRCCKF0EIYoLi4ODg4OiI2NLdbtkd56S1axLV8O5KEpFhERkVHL6+d3sZhqhPQjIQE4cUKus4E2ERFR3jFBMmLHjslBIr28gPLllY6GiIjIcDBBMmJsf0RERFQwTJCMGMc/IiIiKhgmSEbq33/lFCMAEyQiIqL8YoJkpI4eBTQaoGpVwMND6WiIiIgMCxMkI8XqNSIiooJjgmSkmCAREREVHBMkIxQVBVy4INebN1c0FCIiIoPEBMkIHTok/69dG8hlqjoiIiJ6ASZIRojVa0RERK+GCZIRYoJERET0apggGZnwcODGDcDUVE5US0RERPnHBMnIZLQ/qlcPeMEkxURERPQCTJCMTEb1WosWysZBRERkyJggGREh2P6IiIioMDBBMiLXrgH37gFqNdCwodLREBERGS4mSEYkOFj+36gRYGWlbCxERESGjAmSEWH1GhERUeFggmQkNJrMHmxMkIiIiF4NEyQj8fffwKNHgI2N7OJPREREBccEyUhkVK81bQqYmysbCxERkaFjgmQk2P6IiIio8DBBMgJpacCRI3KdA0QSERG9OiZIRiA0FHj6FChVCqhVS+loiIiIDB8TJCOQUb329tuACX+iREREr4wfp0YgY4BItj8iIiIqHEyQDFxSEvDnn3KdCRIREVHhYIJk4E6ckEmSmxtQtarS0RARERkHJkgG7vnu/SqVsrEQEREZCyZIBo7jHxERERW+YpEgLVmyBN7e3rC0tISfnx9OnTqV677NmzeHSqXKtrRr1067T//+/bM937p1a53zPH78GL169YK9vT0cHR0xaNAgxMfH6+0e9SE+Hjh5Uq5z/CMiIqLCo3iCtGnTJgQEBGDatGkICwtDrVq14O/vj+jo6Bz3DwoKwoMHD7TLhQsXYGpqiq5du+rs17p1a539NmzYoPN8r169cPHiRRw4cAA7d+7E0aNHMWTIEL3dpz4cOyYHiSxfHvD2VjoaIiIi46F4gjR//nwMHjwYAwYMwOuvv45ly5bB2toaK1euzHH/0qVLw83NTbscOHAA1tbW2RIktVqts1+pUqW0z12+fBl79+7Fjz/+CD8/PzRp0gSLFi3Cxo0bcf/+fb3eb2Fi9RoREZF+KJogpaSkIDQ0FC1bttRuMzExQcuWLRESEpKnc6xYsQI9evSAjY2NzvbDhw/DxcUFVapUwccff4xHjx5pnwsJCYGjoyPq1q2r3dayZUuYmJjgZEadVRbJycmIi4vTWZTGBImIiEg/FE2QYmJikJ6eDldXV53trq6uiIyMfOnxp06dwoULF/Dhhx/qbG/dujV++uknBAcH48svv8SRI0fQpk0bpKenAwAiIyPh4uKic4yZmRlKly6d63UDAwPh4OCgXTw9PfNzq4Xu8WMgLEyuv/22oqEQEREZHTOlA3gVK1asQI0aNVC/fn2d7T169NCu16hRAzVr1oSPjw8OHz6MFgVszTxx4kQEBARoH8fFxSmaJB05AggBVKsGuLsrFgYREZFRUrQEydnZGaampoiKitLZHhUVBTc3txcem5CQgI0bN2LQoEEvvU6FChXg7OyM69evAwDc3NyyNQJPS0vD48ePc72uWq2Gvb29zqIkVq8RERHpj6IJkoWFBXx9fRGcMZkYAI1Gg+DgYDRs2PCFx27evBnJycno3bv3S69z9+5dPHr0CO7/FbU0bNgQT548QWhoqHaf33//HRqNBn5+fgW8m6LFBImIiEh/FO/FFhAQgOXLl2PNmjW4fPkyPv74YyQkJGDAgAEAgL59+2LixInZjluxYgU6duwIJycnne3x8fEYN24cTpw4gdu3byM4OBjvvfceKlasCH9/fwBAtWrV0Lp1awwePBinTp3Cn3/+ieHDh6NHjx7w8PDQ/02/oshI4NIlOXJ28+ZKR0NERGR8FG+D1L17dzx8+BBTp05FZGQkateujb1792obbkdERMDERDePu3r1Ko4dO4b9+/dnO5+pqSnOnz+PNWvW4MmTJ/Dw8ECrVq0wa9YsqNVq7X7r1q3D8OHD0aJFC5iYmKBLly749ttv9XuzheTQIfl/nTpA6dLKxkJERGSMVEIIoXQQhiguLg4ODg6IjY0t8vZIgwcDP/4IjB0LzJtXpJcmIiIyaHn9/Fa8io3yj+2PiIiI9IsJkoG5fRu4eRMwMwOaNFE6GiIiIuPEBMnAZJQe1a8P2NkpGwsREZGxYoJkYFi9RkREpH9MkAyIEEyQiIiIigITJANy9Srw4AGgVgMvGUeTiIiIXgETJAOSUXrUuDFgaalsLERERMaMCZIByUiQCjjfLhEREeUREyQDodFkjqDN9kdERET6xQTJQJw/Dzx+LLv2162rdDRERETGjQmSgcioXmvaVA4SSURERPrDBMlABAfL/1m9RkREpH9MkAxAaipw9KhcZ4JERESkf0yQDMBffwHx8UDp0kDNmkpHQ0REZPyYIBmAjPZHb78NmPAnRkREpHf8uDUAnF6EiIioaDFBKuaSkoA//5TrHCCSiIioaDBBKuZCQoDkZMDDA6hcWeloiIiISgYmSMXc89VrKpWysRAREZUUTJCKObY/IiIiKnpMkIqxp0+BU6fkOhMkIiKiosMEqRj74w8gLQ2oUAHw8lI6GiIiopKDCVIxxuo1IiIiZTBBKsaYIBERESmDCVIx9egRcPasXH/7bUVDISIiKnGYIBVTR44AQgBvvAG4uSkdDRERUcnCBKmYYvUaERGRcpggFVNMkIiIiJTDBKkYevAAuHxZjpzdrJnS0RAREZU8TJCKoYzSozffBEqVUjYWIiKikogJUjHE6jUiIiJlFYsEacmSJfD29oalpSX8/PxwKmN+jRw0b94cKpUq29KuXTsAQGpqKsaPH48aNWrAxsYGHh4e6Nu3L+7fv69zHm9v72znmDNnjl7vM6+YIBERESlL8QRp06ZNCAgIwLRp0xAWFoZatWrB398f0dHROe4fFBSEBw8eaJcLFy7A1NQUXbt2BQAkJiYiLCwMU6ZMQVhYGIKCgnD16lV06NAh27lmzpypc64RI0bo9V7z4tYt4PZtwMwMaNJE6WiIiIhKJjOlA5g/fz4GDx6MAQMGAACWLVuGXbt2YeXKlZgwYUK2/UuXLq3zeOPGjbC2ttYmSA4ODjhw4IDOPosXL0b9+vURERGBcuXKabfb2dnBrZgNMpRRetSgAWBrq2wsREREJZWiJUgpKSkIDQ1Fy5YttdtMTEzQsmVLhISE5OkcK1asQI8ePWBjY5PrPrGxsVCpVHB0dNTZPmfOHDg5OaFOnTqYN28e0tLScj1HcnIy4uLidBZ9YPUaERGR8hQtQYqJiUF6ejpcXV11tru6uuLKlSsvPf7UqVO4cOECVqxYkes+SUlJGD9+PHr27Al7e3vt9pEjR+LNN99E6dKlcfz4cUycOBEPHjzA/PnzczxPYGAgZsyYkcc7K7inTwETEyZIRERESlK8iu1VrFixAjVq1ED9+vVzfD41NRXdunWDEAJLly7VeS4gIEC7XrNmTVhYWOCjjz5CYGAg1Gp1tnNNnDhR55i4uDh4enoW0p1k2rEDiI0FrK0L/dRERESUR4pWsTk7O8PU1BRRUVE626Oiol7aNighIQEbN27EoEGDcnw+IzkKDw/HgQMHdEqPcuLn54e0tDTcvn07x+fVajXs7e11Fn1xcADMzfV2eiIiInoJRRMkCwsL+Pr6Ijg4WLtNo9EgODgYDRs2fOGxmzdvRnJyMnr37p3tuYzk6Nq1azh48CCcnJxeGsvZs2dhYmICFxeX/N8IERERGRXFq9gCAgLQr18/1K1bF/Xr18fChQuRkJCg7dXWt29flC1bFoGBgTrHrVixAh07dsyW/KSmpuL9999HWFgYdu7cifT0dERGRgKQPeAsLCwQEhKCkydP4u2334adnR1CQkIwZswY9O7dG6U4dDUREVGJp3iC1L17dzx8+BBTp05FZGQkateujb1792obbkdERMDERLeg6+rVqzh27Bj279+f7Xz37t3Djh07AAC1a9fWee7QoUNo3rw51Go1Nm7ciOnTpyM5ORnly5fHmDFjdNoYERERUcmlEkIIpYMwRHFxcXBwcEBsbKxe2yMRERFR4cnr57fiI2kTERERFTdMkIiIiIiyYIJERERElAUTJCIiIqIsmCARERERZcEEiYiIiCgLJkhEREREWTBBIiIiIsqCCRIRERFRFopPNWKoMgYgj4uLUzgSIiIiyquMz+2XTSTCBKmAnj59CgDw9PRUOBIiIiLKr6dPn8LBwSHX5zkXWwFpNBrcv38fdnZ2UKlUSofzyuLi4uDp6Yk7d+6UmLnlSto9836NG+/XuPF+C48QAk+fPoWHhwdMTHJvacQSpAIyMTHBa6+9pnQYhc7e3r5E/PI9r6TdM+/XuPF+jRvvt3C8qOQoAxtpExEREWXBBImIiIgoCyZIBABQq9WYNm0a1Gq10qEUmZJ2z7xf48b7NW6836LHRtpEREREWbAEiYiIiCgLJkhEREREWTBBIiIiIsqCCRIRERFRFkyQSrjAwEDUq1cPdnZ2cHFxQceOHXH16lWlwyoyc+bMgUqlwujRo5UORW/u3buH3r17w8nJCVZWVqhRowb++usvpcPSi/T0dEyZMgXly5eHlZUVfHx8MGvWrJfOuWRIjh49ivbt28PDwwMqlQrbtm3TeV4IgalTp8Ld3R1WVlZo2bIlrl27pkywheBF95uamorx48ejRo0asLGxgYeHB/r27Yv79+8rF/AretnP93lDhw6FSqXCwoULiyy+wpaX+718+TI6dOgABwcH2NjYoF69eoiIiNB7bEyQSrgjR45g2LBhOHHiBA4cOIDU1FS0atUKCQkJSoemd6dPn8b333+PmjVrKh2K3vz7779o3LgxzM3NsWfPHly6dAlff/01SpUqpXRoevHll19i6dKlWLx4MS5fvowvv/wSc+fOxaJFi5QOrdAkJCSgVq1aWLJkSY7Pz507F99++y2WLVuGkydPwsbGBv7+/khKSiriSAvHi+43MTERYWFhmDJlCsLCwhAUFISrV6+iQ4cOCkRaOF72882wdetWnDhxAh4eHkUUmX687H5v3LiBJk2aoGrVqjh8+DDOnz+PKVOmwNLSUv/BCaLnREdHCwDiyJEjSoeiV0+fPhWVKlUSBw4cEM2aNROjRo1SOiS9GD9+vGjSpInSYRSZdu3aiYEDB+ps69y5s+jVq5dCEekXALF161btY41GI9zc3MS8efO02548eSLUarXYsGGDAhEWrqz3m5NTp04JACI8PLxogtKj3O737t27omzZsuLChQvCy8tLLFiwoMhj04ec7rd79+6id+/eisTDEiTSERsbCwAoXbq0wpHo17Bhw9CuXTu0bNlS6VD0aseOHahbty66du0KFxcX1KlTB8uXL1c6LL1p1KgRgoOD8c8//wAAzp07h2PHjqFNmzYKR1Y0bt26hcjISJ33tYODA/z8/BASEqJgZEUnNjYWKpUKjo6OSoeiFxqNBn369MG4cePwxhtvKB2OXmk0GuzatQuVK1eGv78/XFxc4Ofn98Jqx8LEBIm0NBoNRo8ejcaNG6N69epKh6M3GzduRFhYGAIDA5UORe9u3ryJpUuXolKlSti3bx8+/vhjjBw5EmvWrFE6NL2YMGECevTogapVq8Lc3Bx16tTB6NGj0atXL6VDKxKRkZEAAFdXV53trq6u2ueMWVJSEsaPH4+ePXsa7YSuX375JczMzDBy5EilQ9G76OhoxMfHY86cOWjdujX279+PTp06oXPnzjhy5Ijer2+m9yuQwRg2bBguXLiAY8eOKR2K3ty5cwejRo3CgQMHiqYOW2EajQZ169bFF198AQCoU6cOLly4gGXLlqFfv34KR1f4fvnlF6xbtw7r16/HG2+8gbNnz2L06NHw8PAwyvulTKmpqejWrRuEEFi6dKnS4ehFaGgovvnmG4SFhUGlUikdjt5pNBoAwHvvvYcxY8YAAGrXro3jx49j2bJlaNasmV6vzxIkAgAMHz4cO3fuxKFDh/Daa68pHY7ehIaGIjo6Gm+++SbMzMxgZmaGI0eO4Ntvv4WZmRnS09OVDrFQubu74/XXX9fZVq1atSLpAaKEcePGaUuRatSogT59+mDMmDElorQQANzc3AAAUVFROtujoqK0zxmjjOQoPDwcBw4cMNrSoz/++APR0dEoV66c9u9XeHg4Pv30U3h7eysdXqFzdnaGmZmZYn/DWIJUwgkhMGLECGzduhWHDx9G+fLllQ5Jr1q0aIG///5bZ9uAAQNQtWpVjB8/HqampgpFph+NGzfONmzDP//8Ay8vL4Ui0q/ExESYmOh+7zM1NdV+EzV25cuXh5ubG4KDg1G7dm0AQFxcHE6ePImPP/5Y2eD0JCM5unbtGg4dOgQnJyelQ9KbPn36ZGs36e/vjz59+mDAgAEKRaU/FhYWqFevnmJ/w5gglXDDhg3D+vXrsX37dtjZ2WnbKTg4OMDKykrh6AqfnZ1dtvZVNjY2cHJyMsp2V2PGjEGjRo3wxRdfoFu3bjh16hR++OEH/PDDD0qHphft27fH7NmzUa5cObzxxhs4c+YM5s+fj4EDByodWqGJj4/H9evXtY9v3bqFs2fPonTp0ihXrhxGjx6N//u//0OlSpVQvnx5TJkyBR4eHujYsaNyQb+CF92vu7s73n//fYSFhWHnzp1IT0/X/g0rXbo0LCwslAq7wF72882aAJqbm8PNzQ1VqlQp6lALxcvud9y4cejevTuaNm2Kt99+G3v37sVvv/2Gw4cP6z84RfrOUbEBIMdl1apVSodWZIy5m78QQvz222+ievXqQq1Wi6pVq4offvhB6ZD0Ji4uTowaNUqUK1dOWFpaigoVKohJkyaJ5ORkpUMrNIcOHcrxd7Zfv35CCNnVf8qUKcLV1VWo1WrRokULcfXqVWWDfgUvut9bt27l+jfs0KFDSodeIC/7+WZl6N3883K/K1asEBUrVhSWlpaiVq1aYtu2bUUSm0oIIxpiloiIiKgQsJE2ERERURZMkIiIiIiyYIJERERElAUTJCIiIqIsmCARERERZcEEiYiIiCgLJkhEREREWTBBIqISqX///gY7ujQR6R+nGiEio/Oymc6nTZuGb775Bhwnl4hywwSJiIzOgwcPtOubNm3C1KlTdSa8tLW1ha2trRKhEZGBYBUbERkdNzc37eLg4ACVSqWzzdbWNlsVW/PmzTFixAiMHj0apUqVgqurK5YvX46EhAQMGDAAdnZ2qFixIvbs2aNzrQsXLqBNmzawtbWFq6sr+vTpg5iYmCK+YyIqbEyQiIj+s2bNGjg7O+PUqVMYMWIEPv74Y3Tt2hWNGjVCWFgYWrVqhT59+iAxMREA8OTJE7zzzjuoU6cO/vrrL+zduxdRUVHo1q2bwndCRK+KCRIR0X9q1aqFyZMno1KlSpg4cSIsLS3h7OyMwYMHo1KlSpg6dSoePXqE8+fPAwAWL16MOnXq4IsvvkDVqlVRp04drFy5EocOHcI///yj8N0Q0atgGyQiov/UrFlTu25qagonJyfUqFFDu83V1RUAEB0dDQA4d+4cDh06lGN7phs3bqBy5cp6jpiI9IUJEhHRf8zNzXUeq1QqnW0ZveM0Gg0AID4+Hu3bt8eXX36Z7Vzu7u56jJSI9I0JEhFRAb355pvYsmULvL29YWbGP6dExoRtkIiICmjYsGF4/PgxevbsidOnT+PGjRvYt28fBgwYgPT0dKXDI6JXwASJiKiAPDw88OeffyI9PR2tWrVCjRo1MHr0aDg6OsLEhH9eiQyZSnAoWSIiIiId/IpDRERElAUTJCIiIqIsmCARERERZcEEiYiIiCgLJkhEREREWTBBIiIiIsqCCRIRERFRFkyQiIiIiLJggkRERESUBRMkIiIioiyYIBERERFlwQSJiIiIKIv/B9ZBAhW8xt7LAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_training(history.history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.5520 - loss: 0.6732 - val_accuracy: 0.7339 - val_loss: 0.5768\n",
      "Epoch 2/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6873 - loss: 0.5617 - val_accuracy: 0.8283 - val_loss: 0.4466\n",
      "Epoch 3/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 2ms/step - accuracy: 0.7454 - loss: 0.5279 - val_accuracy: 0.8176 - val_loss: 0.4190\n",
      "Epoch 4/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7650 - loss: 0.5017 - val_accuracy: 0.8348 - val_loss: 0.4051\n",
      "Epoch 5/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7930 - loss: 0.4507 - val_accuracy: 0.8262 - val_loss: 0.4038\n",
      "Epoch 6/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7885 - loss: 0.4564 - val_accuracy: 0.8155 - val_loss: 0.3887\n",
      "Epoch 7/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7976 - loss: 0.4555 - val_accuracy: 0.8219 - val_loss: 0.3939\n",
      "Epoch 8/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8125 - loss: 0.4141 - val_accuracy: 0.8262 - val_loss: 0.3924\n",
      "Epoch 9/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 958us/step - accuracy: 0.8250 - loss: 0.4037 - val_accuracy: 0.8262 - val_loss: 0.3944\n",
      "Epoch 10/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 945us/step - accuracy: 0.7960 - loss: 0.4407 - val_accuracy: 0.8155 - val_loss: 0.3962\n",
      "Epoch 11/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8085 - loss: 0.4200 - val_accuracy: 0.8069 - val_loss: 0.3916\n",
      "Epoch 12/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 995us/step - accuracy: 0.8281 - loss: 0.3999 - val_accuracy: 0.8219 - val_loss: 0.3949\n",
      "Epoch 13/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8235 - loss: 0.3940 - val_accuracy: 0.8219 - val_loss: 0.3828\n",
      "Epoch 14/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8165 - loss: 0.3901 - val_accuracy: 0.8197 - val_loss: 0.3931\n",
      "Epoch 15/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8400 - loss: 0.3738 - val_accuracy: 0.8262 - val_loss: 0.3871\n",
      "Epoch 16/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8179 - loss: 0.3934 - val_accuracy: 0.8047 - val_loss: 0.4043\n",
      "Epoch 17/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8340 - loss: 0.4029 - val_accuracy: 0.8004 - val_loss: 0.4027\n",
      "Epoch 18/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 914us/step - accuracy: 0.8297 - loss: 0.3927 - val_accuracy: 0.8240 - val_loss: 0.3883\n",
      "Epoch 19/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8325 - loss: 0.3957 - val_accuracy: 0.8004 - val_loss: 0.4047\n",
      "Epoch 20/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8406 - loss: 0.3817 - val_accuracy: 0.8069 - val_loss: 0.4060\n",
      "Epoch 21/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8344 - loss: 0.3684 - val_accuracy: 0.7961 - val_loss: 0.4106\n",
      "Epoch 22/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8493 - loss: 0.3780 - val_accuracy: 0.7918 - val_loss: 0.4022\n",
      "Epoch 23/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8283 - loss: 0.3681 - val_accuracy: 0.8112 - val_loss: 0.4149\n"
     ]
    }
   ],
   "source": [
    "\"\"\"Keep EarlyStopping and 3 hidden layers.\n",
    "Add dropout with rate of 0.5\n",
    "\"\"\"\n",
    "\n",
    "checkpoint_cb = keras.callbacks.ModelCheckpoint(\n",
    "    'model_checkpoint.keras', save_best_only=True)\n",
    "\n",
    "# Define early stopping callback\n",
    "early_stopping_cb = keras.callbacks.EarlyStopping(\n",
    "    patience=10, restore_best_weights=True)\n",
    "\n",
    "# Define network\n",
    "number_features = X_train_sc.shape[1]\n",
    "model = make_net(number_features, hidden_layers=3, dropout=0.5)\n",
    "\n",
    "### Train model (and store training info in history)\n",
    "history = model.fit(X_train_sc,\n",
    "                    y_train,\n",
    "                    epochs=100,\n",
    "                    batch_size=32,\n",
    "                    validation_data=(X_test_sc, y_test),\n",
    "                    verbose=1,\n",
    "                    callbacks=[checkpoint_cb, early_stopping_cb])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 822us/step\n",
      "\u001b[1m15/15\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 375us/step\n",
      "Training accuracy 0.837\n",
      "Test accuracy 0.822\n"
     ]
    }
   ],
   "source": [
    "calculate_accuracy(model, X_train_sc, X_test_sc, y_train, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.5521 - loss: 0.7286 - val_accuracy: 0.6030 - val_loss: 0.6366\n",
      "Epoch 2/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.5799 - loss: 0.6559 - val_accuracy: 0.6931 - val_loss: 0.5837\n",
      "Epoch 3/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6458 - loss: 0.6387 - val_accuracy: 0.7489 - val_loss: 0.5316\n",
      "Epoch 4/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6782 - loss: 0.5821 - val_accuracy: 0.8090 - val_loss: 0.4784\n",
      "Epoch 5/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7050 - loss: 0.5606 - val_accuracy: 0.7983 - val_loss: 0.4451\n",
      "Epoch 6/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7499 - loss: 0.5106 - val_accuracy: 0.8197 - val_loss: 0.4301\n",
      "Epoch 7/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7868 - loss: 0.4857 - val_accuracy: 0.8112 - val_loss: 0.4158\n",
      "Epoch 8/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7849 - loss: 0.4531 - val_accuracy: 0.8412 - val_loss: 0.3920\n",
      "Epoch 9/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7752 - loss: 0.4861 - val_accuracy: 0.8348 - val_loss: 0.3976\n",
      "Epoch 10/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7822 - loss: 0.4594 - val_accuracy: 0.8348 - val_loss: 0.3914\n",
      "Epoch 11/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 971us/step - accuracy: 0.7885 - loss: 0.4631 - val_accuracy: 0.8326 - val_loss: 0.3998\n",
      "Epoch 12/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7685 - loss: 0.5016 - val_accuracy: 0.8348 - val_loss: 0.3889\n",
      "Epoch 13/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8031 - loss: 0.4424 - val_accuracy: 0.8369 - val_loss: 0.3759\n",
      "Epoch 14/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8167 - loss: 0.4183 - val_accuracy: 0.8476 - val_loss: 0.3722\n",
      "Epoch 15/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8213 - loss: 0.4152 - val_accuracy: 0.8240 - val_loss: 0.3798\n",
      "Epoch 16/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8163 - loss: 0.4082 - val_accuracy: 0.8305 - val_loss: 0.3910\n",
      "Epoch 17/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 942us/step - accuracy: 0.8055 - loss: 0.4163 - val_accuracy: 0.8283 - val_loss: 0.3822\n",
      "Epoch 18/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8019 - loss: 0.4332 - val_accuracy: 0.8369 - val_loss: 0.3796\n",
      "Epoch 19/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 934us/step - accuracy: 0.8201 - loss: 0.4099 - val_accuracy: 0.8412 - val_loss: 0.3780\n",
      "Epoch 20/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 850us/step - accuracy: 0.8199 - loss: 0.3930 - val_accuracy: 0.8219 - val_loss: 0.3830\n",
      "Epoch 21/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8101 - loss: 0.4252 - val_accuracy: 0.8326 - val_loss: 0.3770\n",
      "Epoch 22/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8142 - loss: 0.3924 - val_accuracy: 0.8283 - val_loss: 0.3833\n",
      "Epoch 23/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7955 - loss: 0.4265 - val_accuracy: 0.8326 - val_loss: 0.3724\n",
      "Epoch 24/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 2ms/step - accuracy: 0.8231 - loss: 0.3923 - val_accuracy: 0.8219 - val_loss: 0.3774\n"
     ]
    }
   ],
   "source": [
    "\"\"\"Keep EarlyStopping, 3 hidden layers and 0.5 dropout.\n",
    "Change to 32 hidden layer neurons (from 64)\n",
    "\"\"\"\n",
    "\n",
    "checkpoint_cb = keras.callbacks.ModelCheckpoint(\n",
    "    'model_checkpoint.keras', save_best_only=True)\n",
    "\n",
    "# Define early stopping callback\n",
    "early_stopping_cb = keras.callbacks.EarlyStopping(\n",
    "    patience=10, restore_best_weights=True)\n",
    "\n",
    "# Define network\n",
    "number_features = X_train_sc.shape[1]\n",
    "model = make_net(number_features, hidden_layers=3, dropout=0.5,\n",
    "                 hidden_layer_neurones=32)\n",
    "\n",
    "### Train model (and store training info in history)\n",
    "history = model.fit(X_train_sc,\n",
    "                    y_train,\n",
    "                    epochs=100,\n",
    "                    batch_size=32,\n",
    "                    validation_data=(X_test_sc, y_test),\n",
    "                    verbose=1,\n",
    "                    callbacks=[checkpoint_cb, early_stopping_cb])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 862us/step\n",
      "\u001b[1m15/15\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 720us/step\n",
      "Training accuracy 0.832\n",
      "Test accuracy 0.848\n"
     ]
    }
   ],
   "source": [
    "calculate_accuracy(model, X_train_sc, X_test_sc, y_train, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the following cells, I walk through an example of dropping one of the features (whether the patient has diabetes) from the data as an example of how you would do this.  In practice, you should do this following some analysis via Explainable AI methods, such as the use of Shapley values, which you will learn about in session 4G."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# We cut from X, which already has the label (\"Clotbuster given\") dropped\n",
    "# We can use the same y as before\n",
    "X_cut = X.drop('Diabetes', axis=1)\n",
    "\n",
    "X_np_cut = X_cut.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split data\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    X_np_cut, y_np, test_size = 0.25, random_state=42)\n",
    "\n",
    "# Scale X data\n",
    "X_train_sc, X_test_sc = scale_data(X_train, X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\dan\\anaconda3\\envs\\tf_hsma\\Lib\\site-packages\\keras\\src\\layers\\core\\dense.py:88: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
      "  super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.5255 - loss: 0.7121 - val_accuracy: 0.6009 - val_loss: 0.6663\n",
      "Epoch 2/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 2ms/step - accuracy: 0.6119 - loss: 0.6685 - val_accuracy: 0.6137 - val_loss: 0.6449\n",
      "Epoch 3/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6093 - loss: 0.6471 - val_accuracy: 0.6931 - val_loss: 0.5566\n",
      "Epoch 4/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 2ms/step - accuracy: 0.6685 - loss: 0.5886 - val_accuracy: 0.7854 - val_loss: 0.4961\n",
      "Epoch 5/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.6935 - loss: 0.5392 - val_accuracy: 0.8026 - val_loss: 0.4620\n",
      "Epoch 6/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7225 - loss: 0.5247 - val_accuracy: 0.7918 - val_loss: 0.4429\n",
      "Epoch 7/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 2ms/step - accuracy: 0.7649 - loss: 0.4968 - val_accuracy: 0.8112 - val_loss: 0.4242\n",
      "Epoch 8/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7832 - loss: 0.5054 - val_accuracy: 0.8112 - val_loss: 0.4209\n",
      "Epoch 9/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7763 - loss: 0.4727 - val_accuracy: 0.8197 - val_loss: 0.4113\n",
      "Epoch 10/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8052 - loss: 0.4480 - val_accuracy: 0.8112 - val_loss: 0.4108\n",
      "Epoch 11/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7930 - loss: 0.4660 - val_accuracy: 0.8133 - val_loss: 0.3960\n",
      "Epoch 12/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7972 - loss: 0.4456 - val_accuracy: 0.8197 - val_loss: 0.3940\n",
      "Epoch 13/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7865 - loss: 0.4443 - val_accuracy: 0.8176 - val_loss: 0.4032\n",
      "Epoch 14/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7971 - loss: 0.4593 - val_accuracy: 0.8112 - val_loss: 0.3945\n",
      "Epoch 15/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7964 - loss: 0.4329 - val_accuracy: 0.8090 - val_loss: 0.3985\n",
      "Epoch 16/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7855 - loss: 0.4345 - val_accuracy: 0.8090 - val_loss: 0.3975\n",
      "Epoch 17/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.7984 - loss: 0.4616 - val_accuracy: 0.8240 - val_loss: 0.3913\n",
      "Epoch 18/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8087 - loss: 0.4263 - val_accuracy: 0.8155 - val_loss: 0.3910\n",
      "Epoch 19/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8138 - loss: 0.4199 - val_accuracy: 0.8240 - val_loss: 0.3952\n",
      "Epoch 20/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8190 - loss: 0.4022 - val_accuracy: 0.8197 - val_loss: 0.3943\n",
      "Epoch 21/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8070 - loss: 0.4218 - val_accuracy: 0.8348 - val_loss: 0.3825\n",
      "Epoch 22/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8208 - loss: 0.4084 - val_accuracy: 0.8262 - val_loss: 0.3912\n",
      "Epoch 23/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8273 - loss: 0.3988 - val_accuracy: 0.8155 - val_loss: 0.3992\n",
      "Epoch 24/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8295 - loss: 0.4057 - val_accuracy: 0.8155 - val_loss: 0.3878\n",
      "Epoch 25/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8199 - loss: 0.4170 - val_accuracy: 0.8305 - val_loss: 0.3885\n",
      "Epoch 26/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8262 - loss: 0.4101 - val_accuracy: 0.8133 - val_loss: 0.3868\n",
      "Epoch 27/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8368 - loss: 0.3920 - val_accuracy: 0.8197 - val_loss: 0.3940\n",
      "Epoch 28/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 947us/step - accuracy: 0.8230 - loss: 0.3890 - val_accuracy: 0.8283 - val_loss: 0.3871\n",
      "Epoch 29/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8281 - loss: 0.3892 - val_accuracy: 0.8197 - val_loss: 0.3923\n",
      "Epoch 30/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8222 - loss: 0.4086 - val_accuracy: 0.8090 - val_loss: 0.4054\n",
      "Epoch 31/100\n",
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step - accuracy: 0.8278 - loss: 0.3998 - val_accuracy: 0.8197 - val_loss: 0.3927\n"
     ]
    }
   ],
   "source": [
    "\"\"\"Keep EarlyStopping, 3 hidden layers, 0.5 dropout, and 32 hidden layer\n",
    "neurons.\n",
    "Dropped 'diabetes' feature from data.\n",
    "\"\"\"\n",
    "\n",
    "checkpoint_cb = keras.callbacks.ModelCheckpoint(\n",
    "    'model_checkpoint.keras', save_best_only=True)\n",
    "\n",
    "# Define early stopping callback\n",
    "early_stopping_cb = keras.callbacks.EarlyStopping(\n",
    "    patience=10, restore_best_weights=True)\n",
    "\n",
    "# Define network\n",
    "number_features = X_train_sc.shape[1]\n",
    "model = make_net(number_features, hidden_layers=3, dropout=0.5,\n",
    "                 hidden_layer_neurones=32)\n",
    "\n",
    "### Train model (and store training info in history)\n",
    "history = model.fit(X_train_sc,\n",
    "                    y_train,\n",
    "                    epochs=100,\n",
    "                    batch_size=32,\n",
    "                    validation_data=(X_test_sc, y_test),\n",
    "                    verbose=1,\n",
    "                    callbacks=[checkpoint_cb, early_stopping_cb])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m44/44\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step \n",
      "\u001b[1m15/15\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 364us/step\n",
      "Training accuracy 0.847\n",
      "Test accuracy 0.835\n"
     ]
    }
   ],
   "source": [
    "calculate_accuracy(model, X_train_sc, X_test_sc, y_train, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the above example, dropping the 'diabetes' feature led to a more inaccurate model on the test set."
   ]
  }
 ],
 "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}