All_Features-NNs.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import sklearn\n",
    "import sklearn.preprocessing\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import classification_report, plot_confusion_matrix\n",
    "from sklearn.decomposition import PCA\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import sklearn.pipeline\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from transforms import FourierTransform, MeanTransform"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [],
   "source": [
    "from dataset import HARDatasetCrops"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [],
   "source": [
    "dataset = HARDatasetCrops('motionsense-dataset/train', 256, 50, 50, metadata_file='motionsense-dataset/data_subjects_info.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our dataset is composed of 5205 datapoints, each have shape `(256, 12)` because we have 12 signals of 256 samples each."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [],
   "source": [
    "X_roll = np.array([sample[:,0] for sample, _, _ in dataset])\n",
    "X_pitch = np.array([sample[:,1] for sample, _, _ in dataset])\n",
    "X_yaw = np.array([sample[:,2] for sample, _, _ in dataset])\n",
    "X_gx = np.array([sample[:,3] for sample, _, _ in dataset])\n",
    "X_gy = np.array([sample[:,4] for sample, _, _ in dataset])\n",
    "X_gz = np.array([sample[:,5] for sample, _, _ in dataset])\n",
    "X_acc = np.array([np.linalg.norm(sample[:,6:9], axis=1) for sample, _, _ in dataset])\n",
    "X_rot = np.array([np.linalg.norm(sample[:,9:12], axis=1) for sample, _, _ in dataset])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Classes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In order to fit a `sklearn` model we should encode each class with a unique integer:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [],
   "source": [
    "label_encoder = sklearn.preprocessing.LabelEncoder()\n",
    "label_encoder.fit(list(dataset.CLASSES.keys()));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [],
   "source": [
    "y = label_encoder.transform([cls for _, cls, _ in dataset])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Train-test splitting"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will first \"stack\" our features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4095, 2048)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = np.hstack([X_roll,\n",
    "               X_pitch,\n",
    "               X_yaw,\n",
    "               X_gx,\n",
    "               X_gy,\n",
    "               X_gz,\n",
    "               X_acc,\n",
    "               X_rot])\n",
    "\n",
    "X.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### PCA feature reduction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [],
   "source": [
    "roll_transform = sklearn.pipeline.make_pipeline(\n",
    "    MeanTransform()\n",
    ")\n",
    "pitch_transform = sklearn.pipeline.make_pipeline(\n",
    "    MeanTransform()\n",
    ")\n",
    "yaw_transform = sklearn.pipeline.make_pipeline(\n",
    "    MeanTransform()\n",
    ")\n",
    "gx_transform = sklearn.pipeline.make_pipeline(\n",
    "    MeanTransform()\n",
    ")\n",
    "gy_transform = sklearn.pipeline.make_pipeline(\n",
    "    MeanTransform()\n",
    ")\n",
    "gz_transform = sklearn.pipeline.make_pipeline(\n",
    "    MeanTransform()\n",
    ")\n",
    "acc_transform = sklearn.pipeline.make_pipeline(\n",
    "    FourierTransform(),\n",
    "    PCA(n_components=32)\n",
    ")\n",
    "rot_transform = sklearn.pipeline.make_pipeline(\n",
    "    FourierTransform(),\n",
    "    PCA(n_components=32)\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "metadata = np.array([metadata for _, _, metadata in dataset])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = np.hstack([X, metadata])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### PCA feature reduction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "features = ColumnTransformer([\n",
    "    (\"roll\", roll_transform, slice(0, 256)),\n",
    "    (\"pitch\", pitch_transform, slice(256, 512)),\n",
    "    (\"yaw\", yaw_transform, slice(512, 768)),\n",
    "    (\"gravity-x\", gx_transform, slice(768, 1024)),\n",
    "    (\"gravity-y\", gy_transform, slice(1024, 1280)),\n",
    "    (\"gravity-z\", gz_transform, slice(1280, 1536)),\n",
    "    (\"accelaration\", acc_transform, slice(1536, 1792)),\n",
    "    (\"rotation\", rot_transform, slice(1792, 2048)),\n",
    "    (\"metadata\", \"passthrough\", slice(-4, X.shape[1]))\n",
    "])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Multilayer perceptron"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [],
   "source": [
    "import sklearn.neural_network"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hyperparameter search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import ShuffleSplit\n",
    "from sklearn.model_selection import GridSearchCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "cv = ShuffleSplit(n_splits=5, test_size=0.15, random_state=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf = sklearn.pipeline.make_pipeline(\n",
    "    features,\n",
    "    sklearn.preprocessing.StandardScaler(),\n",
    "    sklearn.neural_network.MLPClassifier()\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid =  {\n",
    "    'mlpclassifier__hidden_layer_sizes': [2, 3, 4],\n",
    "    'mlpclassifier__learning_rate_init': [0.001, 0.0005, 0.0001],\n",
    "    'mlpclassifier__alpha': [0.01, 0.001, 0.0001],\n",
    "    'mlpclassifier__early_stopping': [True],\n",
    "    'mlpclassifier__max_iter': [300],\n",
    "    'mlpclassifier__random_state': [1]\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=ShuffleSplit(n_splits=5, random_state=1, test_size=0.15, train_size=None),\n",
       "             error_score=nan,\n",
       "             estimator=Pipeline(memory=None,\n",
       "                                steps=[('columntransformer',\n",
       "                                        ColumnTransformer(n_jobs=None,\n",
       "                                                          remainder='drop',\n",
       "                                                          sparse_threshold=0.3,\n",
       "                                                          transformer_weights=None,\n",
       "                                                          transformers=[('roll',\n",
       "                                                                         Pipeline(memory=None,\n",
       "                                                                                  steps=[('meantransform',\n",
       "                                                                                          MeanTransform())],\n",
       "                                                                                  verbose=Fa...\n",
       "             iid='deprecated', n_jobs=None,\n",
       "             param_grid={'mlpclassifier__alpha': [0.01, 0.001, 0.0001],\n",
       "                         'mlpclassifier__early_stopping': [True],\n",
       "                         'mlpclassifier__hidden_layer_sizes': [2, 3, 4],\n",
       "                         'mlpclassifier__learning_rate_init': [0.001, 0.0005,\n",
       "                                                               0.0001],\n",
       "                         'mlpclassifier__max_iter': [300],\n",
       "                         'mlpclassifier__random_state': [1]},\n",
       "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
       "             scoring=None, verbose=0)"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search = GridSearchCV(clf, param_grid, cv=cv)\n",
    "\n",
    "grid_search.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'mlpclassifier__alpha': 0.01,\n",
       " 'mlpclassifier__early_stopping': True,\n",
       " 'mlpclassifier__hidden_layer_sizes': 4,\n",
       " 'mlpclassifier__learning_rate_init': 0.001,\n",
       " 'mlpclassifier__max_iter': 300,\n",
       " 'mlpclassifier__random_state': 1}"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9553398058252427"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "cv_results = pd.DataFrame(grid_search.cv_results_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "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>mean_fit_time</th>\n",
       "      <th>std_fit_time</th>\n",
       "      <th>mean_score_time</th>\n",
       "      <th>std_score_time</th>\n",
       "      <th>param_mlpclassifier__alpha</th>\n",
       "      <th>param_mlpclassifier__early_stopping</th>\n",
       "      <th>param_mlpclassifier__hidden_layer_sizes</th>\n",
       "      <th>param_mlpclassifier__learning_rate_init</th>\n",
       "      <th>param_mlpclassifier__max_iter</th>\n",
       "      <th>param_mlpclassifier__random_state</th>\n",
       "      <th>params</th>\n",
       "      <th>split0_test_score</th>\n",
       "      <th>split1_test_score</th>\n",
       "      <th>split2_test_score</th>\n",
       "      <th>split3_test_score</th>\n",
       "      <th>split4_test_score</th>\n",
       "      <th>mean_test_score</th>\n",
       "      <th>std_test_score</th>\n",
       "      <th>rank_test_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.841672</td>\n",
       "      <td>0.171087</td>\n",
       "      <td>0.006134</td>\n",
       "      <td>0.000162</td>\n",
       "      <td>0.01</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "      <td>0.001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.01, 'mlpclassifier_...</td>\n",
       "      <td>0.934466</td>\n",
       "      <td>0.956311</td>\n",
       "      <td>0.963592</td>\n",
       "      <td>0.953883</td>\n",
       "      <td>0.968447</td>\n",
       "      <td>0.955340</td>\n",
       "      <td>0.011650</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.868812</td>\n",
       "      <td>0.123359</td>\n",
       "      <td>0.006737</td>\n",
       "      <td>0.000703</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "      <td>0.001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.0001, 'mlpclassifie...</td>\n",
       "      <td>0.946602</td>\n",
       "      <td>0.956311</td>\n",
       "      <td>0.944175</td>\n",
       "      <td>0.949029</td>\n",
       "      <td>0.946602</td>\n",
       "      <td>0.948544</td>\n",
       "      <td>0.004176</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.869258</td>\n",
       "      <td>0.126058</td>\n",
       "      <td>0.006892</td>\n",
       "      <td>0.000525</td>\n",
       "      <td>0.001</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "      <td>0.001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.001, 'mlpclassifier...</td>\n",
       "      <td>0.949029</td>\n",
       "      <td>0.961165</td>\n",
       "      <td>0.946602</td>\n",
       "      <td>0.944175</td>\n",
       "      <td>0.939320</td>\n",
       "      <td>0.948058</td>\n",
       "      <td>0.007298</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.148775</td>\n",
       "      <td>0.233223</td>\n",
       "      <td>0.006864</td>\n",
       "      <td>0.001073</td>\n",
       "      <td>0.01</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.01, 'mlpclassifier_...</td>\n",
       "      <td>0.832524</td>\n",
       "      <td>0.941748</td>\n",
       "      <td>0.936893</td>\n",
       "      <td>0.817961</td>\n",
       "      <td>0.951456</td>\n",
       "      <td>0.896117</td>\n",
       "      <td>0.058240</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.118824</td>\n",
       "      <td>0.164130</td>\n",
       "      <td>0.006326</td>\n",
       "      <td>0.000198</td>\n",
       "      <td>0.001</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.001, 'mlpclassifier...</td>\n",
       "      <td>0.936893</td>\n",
       "      <td>0.861650</td>\n",
       "      <td>0.847087</td>\n",
       "      <td>0.934466</td>\n",
       "      <td>0.851942</td>\n",
       "      <td>0.886408</td>\n",
       "      <td>0.040510</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.834264</td>\n",
       "      <td>0.216222</td>\n",
       "      <td>0.007405</td>\n",
       "      <td>0.000792</td>\n",
       "      <td>0.01</td>\n",
       "      <td>True</td>\n",
       "      <td>3</td>\n",
       "      <td>0.001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.01, 'mlpclassifier_...</td>\n",
       "      <td>0.859223</td>\n",
       "      <td>0.890777</td>\n",
       "      <td>0.883495</td>\n",
       "      <td>0.898058</td>\n",
       "      <td>0.779126</td>\n",
       "      <td>0.862136</td>\n",
       "      <td>0.043511</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1.111960</td>\n",
       "      <td>0.321720</td>\n",
       "      <td>0.006514</td>\n",
       "      <td>0.000454</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.0001, 'mlpclassifie...</td>\n",
       "      <td>0.936893</td>\n",
       "      <td>0.953883</td>\n",
       "      <td>0.509709</td>\n",
       "      <td>0.856796</td>\n",
       "      <td>0.941748</td>\n",
       "      <td>0.839806</td>\n",
       "      <td>0.168573</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.750092</td>\n",
       "      <td>0.326137</td>\n",
       "      <td>0.006584</td>\n",
       "      <td>0.000254</td>\n",
       "      <td>0.001</td>\n",
       "      <td>True</td>\n",
       "      <td>3</td>\n",
       "      <td>0.001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.001, 'mlpclassifier...</td>\n",
       "      <td>0.766990</td>\n",
       "      <td>0.791262</td>\n",
       "      <td>0.766990</td>\n",
       "      <td>0.929612</td>\n",
       "      <td>0.927184</td>\n",
       "      <td>0.836408</td>\n",
       "      <td>0.075635</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.722252</td>\n",
       "      <td>0.246050</td>\n",
       "      <td>0.006411</td>\n",
       "      <td>0.000312</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>True</td>\n",
       "      <td>3</td>\n",
       "      <td>0.001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.0001, 'mlpclassifie...</td>\n",
       "      <td>0.766990</td>\n",
       "      <td>0.793689</td>\n",
       "      <td>0.893204</td>\n",
       "      <td>0.893204</td>\n",
       "      <td>0.798544</td>\n",
       "      <td>0.829126</td>\n",
       "      <td>0.053411</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.881261</td>\n",
       "      <td>0.190376</td>\n",
       "      <td>0.006790</td>\n",
       "      <td>0.000927</td>\n",
       "      <td>0.001</td>\n",
       "      <td>True</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.001, 'mlpclassifier...</td>\n",
       "      <td>0.766990</td>\n",
       "      <td>0.864078</td>\n",
       "      <td>0.769417</td>\n",
       "      <td>0.771845</td>\n",
       "      <td>0.810680</td>\n",
       "      <td>0.796602</td>\n",
       "      <td>0.037363</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.811515</td>\n",
       "      <td>0.052514</td>\n",
       "      <td>0.006992</td>\n",
       "      <td>0.001047</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>True</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.0001, 'mlpclassifie...</td>\n",
       "      <td>0.771845</td>\n",
       "      <td>0.800971</td>\n",
       "      <td>0.791262</td>\n",
       "      <td>0.783981</td>\n",
       "      <td>0.786408</td>\n",
       "      <td>0.786893</td>\n",
       "      <td>0.009513</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.727538</td>\n",
       "      <td>0.046883</td>\n",
       "      <td>0.006911</td>\n",
       "      <td>0.000930</td>\n",
       "      <td>0.01</td>\n",
       "      <td>True</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.01, 'mlpclassifier_...</td>\n",
       "      <td>0.781553</td>\n",
       "      <td>0.788835</td>\n",
       "      <td>0.796117</td>\n",
       "      <td>0.771845</td>\n",
       "      <td>0.791262</td>\n",
       "      <td>0.785922</td>\n",
       "      <td>0.008464</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.726327</td>\n",
       "      <td>0.325296</td>\n",
       "      <td>0.006358</td>\n",
       "      <td>0.000181</td>\n",
       "      <td>0.01</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "      <td>0.001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.01, 'mlpclassifier_...</td>\n",
       "      <td>0.599515</td>\n",
       "      <td>0.815534</td>\n",
       "      <td>0.473301</td>\n",
       "      <td>0.575243</td>\n",
       "      <td>0.531553</td>\n",
       "      <td>0.599029</td>\n",
       "      <td>0.116448</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.753339</td>\n",
       "      <td>0.279383</td>\n",
       "      <td>0.006888</td>\n",
       "      <td>0.000771</td>\n",
       "      <td>0.001</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "      <td>0.001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.001, 'mlpclassifier...</td>\n",
       "      <td>0.798544</td>\n",
       "      <td>0.604369</td>\n",
       "      <td>0.550971</td>\n",
       "      <td>0.519417</td>\n",
       "      <td>0.514563</td>\n",
       "      <td>0.597573</td>\n",
       "      <td>0.105457</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.709570</td>\n",
       "      <td>0.218611</td>\n",
       "      <td>0.006988</td>\n",
       "      <td>0.000869</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "      <td>0.001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.0001, 'mlpclassifie...</td>\n",
       "      <td>0.485437</td>\n",
       "      <td>0.771845</td>\n",
       "      <td>0.550971</td>\n",
       "      <td>0.587379</td>\n",
       "      <td>0.531553</td>\n",
       "      <td>0.585437</td>\n",
       "      <td>0.098836</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1.100839</td>\n",
       "      <td>0.145060</td>\n",
       "      <td>0.006698</td>\n",
       "      <td>0.000465</td>\n",
       "      <td>0.01</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.01, 'mlpclassifier_...</td>\n",
       "      <td>0.458738</td>\n",
       "      <td>0.412621</td>\n",
       "      <td>0.439320</td>\n",
       "      <td>0.458738</td>\n",
       "      <td>0.371359</td>\n",
       "      <td>0.428155</td>\n",
       "      <td>0.033053</td>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1.031099</td>\n",
       "      <td>0.198923</td>\n",
       "      <td>0.006514</td>\n",
       "      <td>0.000379</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.0001, 'mlpclassifie...</td>\n",
       "      <td>0.463592</td>\n",
       "      <td>0.371359</td>\n",
       "      <td>0.388350</td>\n",
       "      <td>0.429612</td>\n",
       "      <td>0.405340</td>\n",
       "      <td>0.411650</td>\n",
       "      <td>0.032332</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1.108589</td>\n",
       "      <td>0.587305</td>\n",
       "      <td>0.007326</td>\n",
       "      <td>0.000489</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>True</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.0001, 'mlpclassifie...</td>\n",
       "      <td>0.308252</td>\n",
       "      <td>0.388350</td>\n",
       "      <td>0.266990</td>\n",
       "      <td>0.716019</td>\n",
       "      <td>0.281553</td>\n",
       "      <td>0.392233</td>\n",
       "      <td>0.167237</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.915267</td>\n",
       "      <td>0.414667</td>\n",
       "      <td>0.006205</td>\n",
       "      <td>0.000126</td>\n",
       "      <td>0.001</td>\n",
       "      <td>True</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.001, 'mlpclassifier...</td>\n",
       "      <td>0.422330</td>\n",
       "      <td>0.383495</td>\n",
       "      <td>0.415049</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.150485</td>\n",
       "      <td>0.374272</td>\n",
       "      <td>0.118275</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.943183</td>\n",
       "      <td>0.653232</td>\n",
       "      <td>0.007047</td>\n",
       "      <td>0.000983</td>\n",
       "      <td>0.001</td>\n",
       "      <td>True</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.001, 'mlpclassifier...</td>\n",
       "      <td>0.393204</td>\n",
       "      <td>0.708738</td>\n",
       "      <td>0.264563</td>\n",
       "      <td>0.135922</td>\n",
       "      <td>0.281553</td>\n",
       "      <td>0.356796</td>\n",
       "      <td>0.193980</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.552397</td>\n",
       "      <td>0.174409</td>\n",
       "      <td>0.006343</td>\n",
       "      <td>0.000227</td>\n",
       "      <td>0.01</td>\n",
       "      <td>True</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.01, 'mlpclassifier_...</td>\n",
       "      <td>0.317961</td>\n",
       "      <td>0.286408</td>\n",
       "      <td>0.281553</td>\n",
       "      <td>0.138350</td>\n",
       "      <td>0.286408</td>\n",
       "      <td>0.262136</td>\n",
       "      <td>0.063237</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.358474</td>\n",
       "      <td>0.237928</td>\n",
       "      <td>0.006782</td>\n",
       "      <td>0.000479</td>\n",
       "      <td>0.01</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.01, 'mlpclassifier_...</td>\n",
       "      <td>0.082524</td>\n",
       "      <td>0.055825</td>\n",
       "      <td>0.077670</td>\n",
       "      <td>0.514563</td>\n",
       "      <td>0.075243</td>\n",
       "      <td>0.161165</td>\n",
       "      <td>0.176932</td>\n",
       "      <td>22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.317378</td>\n",
       "      <td>0.225220</td>\n",
       "      <td>0.006588</td>\n",
       "      <td>0.000613</td>\n",
       "      <td>0.001</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.001, 'mlpclassifier...</td>\n",
       "      <td>0.082524</td>\n",
       "      <td>0.053398</td>\n",
       "      <td>0.451456</td>\n",
       "      <td>0.075243</td>\n",
       "      <td>0.075243</td>\n",
       "      <td>0.147573</td>\n",
       "      <td>0.152255</td>\n",
       "      <td>23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.207541</td>\n",
       "      <td>0.007487</td>\n",
       "      <td>0.006912</td>\n",
       "      <td>0.000930</td>\n",
       "      <td>0.001</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.001, 'mlpclassifier...</td>\n",
       "      <td>0.082524</td>\n",
       "      <td>0.063107</td>\n",
       "      <td>0.080097</td>\n",
       "      <td>0.082524</td>\n",
       "      <td>0.082524</td>\n",
       "      <td>0.078155</td>\n",
       "      <td>0.007583</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.206026</td>\n",
       "      <td>0.004940</td>\n",
       "      <td>0.006577</td>\n",
       "      <td>0.000803</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.0001, 'mlpclassifie...</td>\n",
       "      <td>0.087379</td>\n",
       "      <td>0.063107</td>\n",
       "      <td>0.075243</td>\n",
       "      <td>0.082524</td>\n",
       "      <td>0.075243</td>\n",
       "      <td>0.076699</td>\n",
       "      <td>0.008209</td>\n",
       "      <td>25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.203786</td>\n",
       "      <td>0.002695</td>\n",
       "      <td>0.006179</td>\n",
       "      <td>0.000051</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.0001, 'mlpclassifie...</td>\n",
       "      <td>0.087379</td>\n",
       "      <td>0.060680</td>\n",
       "      <td>0.075243</td>\n",
       "      <td>0.082524</td>\n",
       "      <td>0.072816</td>\n",
       "      <td>0.075728</td>\n",
       "      <td>0.009133</td>\n",
       "      <td>26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.218469</td>\n",
       "      <td>0.011922</td>\n",
       "      <td>0.007432</td>\n",
       "      <td>0.001089</td>\n",
       "      <td>0.01</td>\n",
       "      <td>True</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>300</td>\n",
       "      <td>1</td>\n",
       "      <td>{'mlpclassifier__alpha': 0.01, 'mlpclassifier_...</td>\n",
       "      <td>0.084951</td>\n",
       "      <td>0.060680</td>\n",
       "      <td>0.072816</td>\n",
       "      <td>0.080097</td>\n",
       "      <td>0.075243</td>\n",
       "      <td>0.074757</td>\n",
       "      <td>0.008181</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    mean_fit_time  std_fit_time  mean_score_time  std_score_time  \\\n",
       "6        0.841672      0.171087         0.006134        0.000162   \n",
       "24       0.868812      0.123359         0.006737        0.000703   \n",
       "15       0.869258      0.126058         0.006892        0.000525   \n",
       "7        1.148775      0.233223         0.006864        0.001073   \n",
       "16       1.118824      0.164130         0.006326        0.000198   \n",
       "3        0.834264      0.216222         0.007405        0.000792   \n",
       "25       1.111960      0.321720         0.006514        0.000454   \n",
       "12       0.750092      0.326137         0.006584        0.000254   \n",
       "21       0.722252      0.246050         0.006411        0.000312   \n",
       "13       0.881261      0.190376         0.006790        0.000927   \n",
       "22       0.811515      0.052514         0.006992        0.001047   \n",
       "4        0.727538      0.046883         0.006911        0.000930   \n",
       "0        0.726327      0.325296         0.006358        0.000181   \n",
       "9        0.753339      0.279383         0.006888        0.000771   \n",
       "18       0.709570      0.218611         0.006988        0.000869   \n",
       "8        1.100839      0.145060         0.006698        0.000465   \n",
       "26       1.031099      0.198923         0.006514        0.000379   \n",
       "23       1.108589      0.587305         0.007326        0.000489   \n",
       "17       0.915267      0.414667         0.006205        0.000126   \n",
       "14       0.943183      0.653232         0.007047        0.000983   \n",
       "5        0.552397      0.174409         0.006343        0.000227   \n",
       "1        0.358474      0.237928         0.006782        0.000479   \n",
       "10       0.317378      0.225220         0.006588        0.000613   \n",
       "11       0.207541      0.007487         0.006912        0.000930   \n",
       "19       0.206026      0.004940         0.006577        0.000803   \n",
       "20       0.203786      0.002695         0.006179        0.000051   \n",
       "2        0.218469      0.011922         0.007432        0.001089   \n",
       "\n",
       "   param_mlpclassifier__alpha param_mlpclassifier__early_stopping  \\\n",
       "6                        0.01                                True   \n",
       "24                     0.0001                                True   \n",
       "15                      0.001                                True   \n",
       "7                        0.01                                True   \n",
       "16                      0.001                                True   \n",
       "3                        0.01                                True   \n",
       "25                     0.0001                                True   \n",
       "12                      0.001                                True   \n",
       "21                     0.0001                                True   \n",
       "13                      0.001                                True   \n",
       "22                     0.0001                                True   \n",
       "4                        0.01                                True   \n",
       "0                        0.01                                True   \n",
       "9                       0.001                                True   \n",
       "18                     0.0001                                True   \n",
       "8                        0.01                                True   \n",
       "26                     0.0001                                True   \n",
       "23                     0.0001                                True   \n",
       "17                      0.001                                True   \n",
       "14                      0.001                                True   \n",
       "5                        0.01                                True   \n",
       "1                        0.01                                True   \n",
       "10                      0.001                                True   \n",
       "11                      0.001                                True   \n",
       "19                     0.0001                                True   \n",
       "20                     0.0001                                True   \n",
       "2                        0.01                                True   \n",
       "\n",
       "   param_mlpclassifier__hidden_layer_sizes  \\\n",
       "6                                        4   \n",
       "24                                       4   \n",
       "15                                       4   \n",
       "7                                        4   \n",
       "16                                       4   \n",
       "3                                        3   \n",
       "25                                       4   \n",
       "12                                       3   \n",
       "21                                       3   \n",
       "13                                       3   \n",
       "22                                       3   \n",
       "4                                        3   \n",
       "0                                        2   \n",
       "9                                        2   \n",
       "18                                       2   \n",
       "8                                        4   \n",
       "26                                       4   \n",
       "23                                       3   \n",
       "17                                       4   \n",
       "14                                       3   \n",
       "5                                        3   \n",
       "1                                        2   \n",
       "10                                       2   \n",
       "11                                       2   \n",
       "19                                       2   \n",
       "20                                       2   \n",
       "2                                        2   \n",
       "\n",
       "   param_mlpclassifier__learning_rate_init param_mlpclassifier__max_iter  \\\n",
       "6                                    0.001                           300   \n",
       "24                                   0.001                           300   \n",
       "15                                   0.001                           300   \n",
       "7                                   0.0005                           300   \n",
       "16                                  0.0005                           300   \n",
       "3                                    0.001                           300   \n",
       "25                                  0.0005                           300   \n",
       "12                                   0.001                           300   \n",
       "21                                   0.001                           300   \n",
       "13                                  0.0005                           300   \n",
       "22                                  0.0005                           300   \n",
       "4                                   0.0005                           300   \n",
       "0                                    0.001                           300   \n",
       "9                                    0.001                           300   \n",
       "18                                   0.001                           300   \n",
       "8                                   0.0001                           300   \n",
       "26                                  0.0001                           300   \n",
       "23                                  0.0001                           300   \n",
       "17                                  0.0001                           300   \n",
       "14                                  0.0001                           300   \n",
       "5                                   0.0001                           300   \n",
       "1                                   0.0005                           300   \n",
       "10                                  0.0005                           300   \n",
       "11                                  0.0001                           300   \n",
       "19                                  0.0005                           300   \n",
       "20                                  0.0001                           300   \n",
       "2                                   0.0001                           300   \n",
       "\n",
       "   param_mlpclassifier__random_state  \\\n",
       "6                                  1   \n",
       "24                                 1   \n",
       "15                                 1   \n",
       "7                                  1   \n",
       "16                                 1   \n",
       "3                                  1   \n",
       "25                                 1   \n",
       "12                                 1   \n",
       "21                                 1   \n",
       "13                                 1   \n",
       "22                                 1   \n",
       "4                                  1   \n",
       "0                                  1   \n",
       "9                                  1   \n",
       "18                                 1   \n",
       "8                                  1   \n",
       "26                                 1   \n",
       "23                                 1   \n",
       "17                                 1   \n",
       "14                                 1   \n",
       "5                                  1   \n",
       "1                                  1   \n",
       "10                                 1   \n",
       "11                                 1   \n",
       "19                                 1   \n",
       "20                                 1   \n",
       "2                                  1   \n",
       "\n",
       "                                               params  split0_test_score  \\\n",
       "6   {'mlpclassifier__alpha': 0.01, 'mlpclassifier_...           0.934466   \n",
       "24  {'mlpclassifier__alpha': 0.0001, 'mlpclassifie...           0.946602   \n",
       "15  {'mlpclassifier__alpha': 0.001, 'mlpclassifier...           0.949029   \n",
       "7   {'mlpclassifier__alpha': 0.01, 'mlpclassifier_...           0.832524   \n",
       "16  {'mlpclassifier__alpha': 0.001, 'mlpclassifier...           0.936893   \n",
       "3   {'mlpclassifier__alpha': 0.01, 'mlpclassifier_...           0.859223   \n",
       "25  {'mlpclassifier__alpha': 0.0001, 'mlpclassifie...           0.936893   \n",
       "12  {'mlpclassifier__alpha': 0.001, 'mlpclassifier...           0.766990   \n",
       "21  {'mlpclassifier__alpha': 0.0001, 'mlpclassifie...           0.766990   \n",
       "13  {'mlpclassifier__alpha': 0.001, 'mlpclassifier...           0.766990   \n",
       "22  {'mlpclassifier__alpha': 0.0001, 'mlpclassifie...           0.771845   \n",
       "4   {'mlpclassifier__alpha': 0.01, 'mlpclassifier_...           0.781553   \n",
       "0   {'mlpclassifier__alpha': 0.01, 'mlpclassifier_...           0.599515   \n",
       "9   {'mlpclassifier__alpha': 0.001, 'mlpclassifier...           0.798544   \n",
       "18  {'mlpclassifier__alpha': 0.0001, 'mlpclassifie...           0.485437   \n",
       "8   {'mlpclassifier__alpha': 0.01, 'mlpclassifier_...           0.458738   \n",
       "26  {'mlpclassifier__alpha': 0.0001, 'mlpclassifie...           0.463592   \n",
       "23  {'mlpclassifier__alpha': 0.0001, 'mlpclassifie...           0.308252   \n",
       "17  {'mlpclassifier__alpha': 0.001, 'mlpclassifier...           0.422330   \n",
       "14  {'mlpclassifier__alpha': 0.001, 'mlpclassifier...           0.393204   \n",
       "5   {'mlpclassifier__alpha': 0.01, 'mlpclassifier_...           0.317961   \n",
       "1   {'mlpclassifier__alpha': 0.01, 'mlpclassifier_...           0.082524   \n",
       "10  {'mlpclassifier__alpha': 0.001, 'mlpclassifier...           0.082524   \n",
       "11  {'mlpclassifier__alpha': 0.001, 'mlpclassifier...           0.082524   \n",
       "19  {'mlpclassifier__alpha': 0.0001, 'mlpclassifie...           0.087379   \n",
       "20  {'mlpclassifier__alpha': 0.0001, 'mlpclassifie...           0.087379   \n",
       "2   {'mlpclassifier__alpha': 0.01, 'mlpclassifier_...           0.084951   \n",
       "\n",
       "    split1_test_score  split2_test_score  split3_test_score  \\\n",
       "6            0.956311           0.963592           0.953883   \n",
       "24           0.956311           0.944175           0.949029   \n",
       "15           0.961165           0.946602           0.944175   \n",
       "7            0.941748           0.936893           0.817961   \n",
       "16           0.861650           0.847087           0.934466   \n",
       "3            0.890777           0.883495           0.898058   \n",
       "25           0.953883           0.509709           0.856796   \n",
       "12           0.791262           0.766990           0.929612   \n",
       "21           0.793689           0.893204           0.893204   \n",
       "13           0.864078           0.769417           0.771845   \n",
       "22           0.800971           0.791262           0.783981   \n",
       "4            0.788835           0.796117           0.771845   \n",
       "0            0.815534           0.473301           0.575243   \n",
       "9            0.604369           0.550971           0.519417   \n",
       "18           0.771845           0.550971           0.587379   \n",
       "8            0.412621           0.439320           0.458738   \n",
       "26           0.371359           0.388350           0.429612   \n",
       "23           0.388350           0.266990           0.716019   \n",
       "17           0.383495           0.415049           0.500000   \n",
       "14           0.708738           0.264563           0.135922   \n",
       "5            0.286408           0.281553           0.138350   \n",
       "1            0.055825           0.077670           0.514563   \n",
       "10           0.053398           0.451456           0.075243   \n",
       "11           0.063107           0.080097           0.082524   \n",
       "19           0.063107           0.075243           0.082524   \n",
       "20           0.060680           0.075243           0.082524   \n",
       "2            0.060680           0.072816           0.080097   \n",
       "\n",
       "    split4_test_score  mean_test_score  std_test_score  rank_test_score  \n",
       "6            0.968447         0.955340        0.011650                1  \n",
       "24           0.946602         0.948544        0.004176                2  \n",
       "15           0.939320         0.948058        0.007298                3  \n",
       "7            0.951456         0.896117        0.058240                4  \n",
       "16           0.851942         0.886408        0.040510                5  \n",
       "3            0.779126         0.862136        0.043511                6  \n",
       "25           0.941748         0.839806        0.168573                7  \n",
       "12           0.927184         0.836408        0.075635                8  \n",
       "21           0.798544         0.829126        0.053411                9  \n",
       "13           0.810680         0.796602        0.037363               10  \n",
       "22           0.786408         0.786893        0.009513               11  \n",
       "4            0.791262         0.785922        0.008464               12  \n",
       "0            0.531553         0.599029        0.116448               13  \n",
       "9            0.514563         0.597573        0.105457               14  \n",
       "18           0.531553         0.585437        0.098836               15  \n",
       "8            0.371359         0.428155        0.033053               16  \n",
       "26           0.405340         0.411650        0.032332               17  \n",
       "23           0.281553         0.392233        0.167237               18  \n",
       "17           0.150485         0.374272        0.118275               19  \n",
       "14           0.281553         0.356796        0.193980               20  \n",
       "5            0.286408         0.262136        0.063237               21  \n",
       "1            0.075243         0.161165        0.176932               22  \n",
       "10           0.075243         0.147573        0.152255               23  \n",
       "11           0.082524         0.078155        0.007583               24  \n",
       "19           0.075243         0.076699        0.008209               25  \n",
       "20           0.072816         0.075728        0.009133               26  \n",
       "2            0.075243         0.074757        0.008181               27  "
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results.sort_values(\"rank_test_score\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         dws       0.87      0.78      0.82       236\n",
      "         jog       0.97      0.91      0.94       265\n",
      "         sit       0.99      1.00      0.99       663\n",
      "         std       0.93      1.00      0.96       615\n",
      "         ups       0.93      0.85      0.89       298\n",
      "         wlk       0.95      0.98      0.97       666\n",
      "\n",
      "    accuracy                           0.95      2743\n",
      "   macro avg       0.94      0.92      0.93      2743\n",
      "weighted avg       0.95      0.95      0.95      2743\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_train, grid_search.predict(X_train), target_names=label_encoder.classes_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x139f595d0>"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(grid_search, X_train, y_train,\n",
    "                      display_labels=label_encoder.classes_,\n",
    "                      cmap=plt.cm.Blues,\n",
    "                      normalize=\"true\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         dws       0.88      0.64      0.74       123\n",
      "         jog       0.96      0.92      0.94       119\n",
      "         sit       0.99      0.99      0.99       341\n",
      "         std       0.89      1.00      0.94       285\n",
      "         ups       0.91      0.86      0.88       138\n",
      "         wlk       0.94      0.98      0.96       346\n",
      "\n",
      "    accuracy                           0.94      1352\n",
      "   macro avg       0.93      0.90      0.91      1352\n",
      "weighted avg       0.94      0.94      0.94      1352\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, grid_search.predict(X_test), target_names=label_encoder.classes_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x13a952690>"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(grid_search, X_test, y_test,\n",
    "                      display_labels=label_encoder.classes_,\n",
    "                      cmap=plt.cm.Blues,\n",
    "                      normalize=\"true\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Class-balancing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "import imblearn.pipeline\n",
    "import imblearn.over_sampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf = imblearn.pipeline.Pipeline([\n",
    "    ('features', features),\n",
    "    ('balancing', imblearn.over_sampling.SMOTE(random_state=1)),\n",
    "    ('normalize', sklearn.preprocessing.StandardScaler()),\n",
    "    ('model', sklearn.neural_network.MLPClassifier(early_stopping=True, max_iter=300, random_state=1))\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid =  {\n",
    "    'model__hidden_layer_sizes': [2, 3, 4],\n",
    "    'model__learning_rate_init': [0.001, 0.0005, 0.0001],\n",
    "    'model__alpha': [0.01, 0.001, 0.0001]\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/sergio/PycharmProjects/MLEARN/venv/lib/python3.7/site-packages/sklearn/neural_network/_multilayer_perceptron.py:571: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (300) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/Users/sergio/PycharmProjects/MLEARN/venv/lib/python3.7/site-packages/sklearn/neural_network/_multilayer_perceptron.py:571: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (300) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/Users/sergio/PycharmProjects/MLEARN/venv/lib/python3.7/site-packages/sklearn/neural_network/_multilayer_perceptron.py:571: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (300) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/Users/sergio/PycharmProjects/MLEARN/venv/lib/python3.7/site-packages/sklearn/neural_network/_multilayer_perceptron.py:571: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (300) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/Users/sergio/PycharmProjects/MLEARN/venv/lib/python3.7/site-packages/sklearn/neural_network/_multilayer_perceptron.py:571: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (300) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/Users/sergio/PycharmProjects/MLEARN/venv/lib/python3.7/site-packages/sklearn/neural_network/_multilayer_perceptron.py:571: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (300) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n",
      "/Users/sergio/PycharmProjects/MLEARN/venv/lib/python3.7/site-packages/sklearn/neural_network/_multilayer_perceptron.py:571: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (300) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=ShuffleSplit(n_splits=5, random_state=1, test_size=0.15, train_size=None),\n",
       "             error_score=nan,\n",
       "             estimator=Pipeline(memory=None,\n",
       "                                steps=[('features',\n",
       "                                        ColumnTransformer(n_jobs=None,\n",
       "                                                          remainder='drop',\n",
       "                                                          sparse_threshold=0.3,\n",
       "                                                          transformer_weights=None,\n",
       "                                                          transformers=[('roll',\n",
       "                                                                         Pipeline(memory=None,\n",
       "                                                                                  steps=[('meantransform',\n",
       "                                                                                          MeanTransform())],\n",
       "                                                                                  verbose=False),\n",
       "                                                                         slic...\n",
       "                                                      shuffle=True,\n",
       "                                                      solver='adam', tol=0.0001,\n",
       "                                                      validation_fraction=0.1,\n",
       "                                                      verbose=False,\n",
       "                                                      warm_start=False))],\n",
       "                                verbose=False),\n",
       "             iid='deprecated', n_jobs=None,\n",
       "             param_grid={'model__alpha': [0.01, 0.001, 0.0001],\n",
       "                         'model__hidden_layer_sizes': [2, 3, 4],\n",
       "                         'model__learning_rate_init': [0.001, 0.0005, 0.0001]},\n",
       "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
       "             scoring=None, verbose=0)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search = GridSearchCV(clf, param_grid, cv=cv)\n",
    "\n",
    "grid_search.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'model__alpha': 0.0001,\n",
       " 'model__hidden_layer_sizes': 4,\n",
       " 'model__learning_rate_init': 0.001}"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9621359223300973"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "cv_results = pd.DataFrame(grid_search.cv_results_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "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>mean_fit_time</th>\n",
       "      <th>std_fit_time</th>\n",
       "      <th>mean_score_time</th>\n",
       "      <th>std_score_time</th>\n",
       "      <th>param_model__alpha</th>\n",
       "      <th>param_model__hidden_layer_sizes</th>\n",
       "      <th>param_model__learning_rate_init</th>\n",
       "      <th>params</th>\n",
       "      <th>split0_test_score</th>\n",
       "      <th>split1_test_score</th>\n",
       "      <th>split2_test_score</th>\n",
       "      <th>split3_test_score</th>\n",
       "      <th>split4_test_score</th>\n",
       "      <th>mean_test_score</th>\n",
       "      <th>std_test_score</th>\n",
       "      <th>rank_test_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.970589</td>\n",
       "      <td>0.158883</td>\n",
       "      <td>0.006192</td>\n",
       "      <td>0.000285</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>4</td>\n",
       "      <td>0.001</td>\n",
       "      <td>{'model__alpha': 0.0001, 'model__hidden_layer_...</td>\n",
       "      <td>0.953883</td>\n",
       "      <td>0.966019</td>\n",
       "      <td>0.951456</td>\n",
       "      <td>0.966019</td>\n",
       "      <td>0.973301</td>\n",
       "      <td>0.962136</td>\n",
       "      <td>0.008209</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.043675</td>\n",
       "      <td>0.139572</td>\n",
       "      <td>0.006736</td>\n",
       "      <td>0.000917</td>\n",
       "      <td>0.01</td>\n",
       "      <td>4</td>\n",
       "      <td>0.001</td>\n",
       "      <td>{'model__alpha': 0.01, 'model__hidden_layer_si...</td>\n",
       "      <td>0.956311</td>\n",
       "      <td>0.966019</td>\n",
       "      <td>0.946602</td>\n",
       "      <td>0.961165</td>\n",
       "      <td>0.978155</td>\n",
       "      <td>0.961650</td>\n",
       "      <td>0.010457</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.918648</td>\n",
       "      <td>0.134335</td>\n",
       "      <td>0.006069</td>\n",
       "      <td>0.000223</td>\n",
       "      <td>0.001</td>\n",
       "      <td>4</td>\n",
       "      <td>0.001</td>\n",
       "      <td>{'model__alpha': 0.001, 'model__hidden_layer_s...</td>\n",
       "      <td>0.941748</td>\n",
       "      <td>0.966019</td>\n",
       "      <td>0.946602</td>\n",
       "      <td>0.961165</td>\n",
       "      <td>0.978155</td>\n",
       "      <td>0.958738</td>\n",
       "      <td>0.013205</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.322700</td>\n",
       "      <td>0.160295</td>\n",
       "      <td>0.006105</td>\n",
       "      <td>0.000126</td>\n",
       "      <td>0.001</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>{'model__alpha': 0.001, 'model__hidden_layer_s...</td>\n",
       "      <td>0.949029</td>\n",
       "      <td>0.963592</td>\n",
       "      <td>0.941748</td>\n",
       "      <td>0.951456</td>\n",
       "      <td>0.956311</td>\n",
       "      <td>0.952427</td>\n",
       "      <td>0.007298</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.277122</td>\n",
       "      <td>0.127310</td>\n",
       "      <td>0.006284</td>\n",
       "      <td>0.000298</td>\n",
       "      <td>0.01</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>{'model__alpha': 0.01, 'model__hidden_layer_si...</td>\n",
       "      <td>0.949029</td>\n",
       "      <td>0.958738</td>\n",
       "      <td>0.932039</td>\n",
       "      <td>0.946602</td>\n",
       "      <td>0.970874</td>\n",
       "      <td>0.951456</td>\n",
       "      <td>0.012935</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1.227781</td>\n",
       "      <td>0.227956</td>\n",
       "      <td>0.006047</td>\n",
       "      <td>0.000276</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>{'model__alpha': 0.0001, 'model__hidden_layer_...</td>\n",
       "      <td>0.898058</td>\n",
       "      <td>0.958738</td>\n",
       "      <td>0.941748</td>\n",
       "      <td>0.958738</td>\n",
       "      <td>0.966019</td>\n",
       "      <td>0.944660</td>\n",
       "      <td>0.024628</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1.112600</td>\n",
       "      <td>0.095687</td>\n",
       "      <td>0.006063</td>\n",
       "      <td>0.000236</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3</td>\n",
       "      <td>0.001</td>\n",
       "      <td>{'model__alpha': 0.001, 'model__hidden_layer_s...</td>\n",
       "      <td>0.919903</td>\n",
       "      <td>0.941748</td>\n",
       "      <td>0.944175</td>\n",
       "      <td>0.956311</td>\n",
       "      <td>0.961165</td>\n",
       "      <td>0.944660</td>\n",
       "      <td>0.014351</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1.118584</td>\n",
       "      <td>0.034868</td>\n",
       "      <td>0.006000</td>\n",
       "      <td>0.000193</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>3</td>\n",
       "      <td>0.001</td>\n",
       "      <td>{'model__alpha': 0.0001, 'model__hidden_layer_...</td>\n",
       "      <td>0.922330</td>\n",
       "      <td>0.932039</td>\n",
       "      <td>0.936893</td>\n",
       "      <td>0.953883</td>\n",
       "      <td>0.953883</td>\n",
       "      <td>0.939806</td>\n",
       "      <td>0.012414</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1.810017</td>\n",
       "      <td>0.288030</td>\n",
       "      <td>0.006075</td>\n",
       "      <td>0.000220</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>{'model__alpha': 0.001, 'model__hidden_layer_s...</td>\n",
       "      <td>0.922330</td>\n",
       "      <td>0.905340</td>\n",
       "      <td>0.946602</td>\n",
       "      <td>0.953883</td>\n",
       "      <td>0.939320</td>\n",
       "      <td>0.933495</td>\n",
       "      <td>0.017543</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.252790</td>\n",
       "      <td>0.393719</td>\n",
       "      <td>0.006880</td>\n",
       "      <td>0.000538</td>\n",
       "      <td>0.01</td>\n",
       "      <td>3</td>\n",
       "      <td>0.001</td>\n",
       "      <td>{'model__alpha': 0.01, 'model__hidden_layer_si...</td>\n",
       "      <td>0.946602</td>\n",
       "      <td>0.915049</td>\n",
       "      <td>0.951456</td>\n",
       "      <td>0.885922</td>\n",
       "      <td>0.963592</td>\n",
       "      <td>0.932524</td>\n",
       "      <td>0.028281</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.724635</td>\n",
       "      <td>0.244262</td>\n",
       "      <td>0.006367</td>\n",
       "      <td>0.000545</td>\n",
       "      <td>0.01</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>{'model__alpha': 0.01, 'model__hidden_layer_si...</td>\n",
       "      <td>0.917476</td>\n",
       "      <td>0.917476</td>\n",
       "      <td>0.927184</td>\n",
       "      <td>0.929612</td>\n",
       "      <td>0.958738</td>\n",
       "      <td>0.930097</td>\n",
       "      <td>0.015150</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1.442220</td>\n",
       "      <td>0.130967</td>\n",
       "      <td>0.006104</td>\n",
       "      <td>0.000220</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>{'model__alpha': 0.0001, 'model__hidden_layer_...</td>\n",
       "      <td>0.895631</td>\n",
       "      <td>0.890777</td>\n",
       "      <td>0.932039</td>\n",
       "      <td>0.927184</td>\n",
       "      <td>0.934466</td>\n",
       "      <td>0.916019</td>\n",
       "      <td>0.018838</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.733813</td>\n",
       "      <td>0.297156</td>\n",
       "      <td>0.007035</td>\n",
       "      <td>0.000360</td>\n",
       "      <td>0.01</td>\n",
       "      <td>2</td>\n",
       "      <td>0.001</td>\n",
       "      <td>{'model__alpha': 0.01, 'model__hidden_layer_si...</td>\n",
       "      <td>0.873786</td>\n",
       "      <td>0.837379</td>\n",
       "      <td>0.837379</td>\n",
       "      <td>0.854369</td>\n",
       "      <td>0.895631</td>\n",
       "      <td>0.859709</td>\n",
       "      <td>0.022425</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1.496390</td>\n",
       "      <td>0.479643</td>\n",
       "      <td>0.006340</td>\n",
       "      <td>0.000201</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>2</td>\n",
       "      <td>0.001</td>\n",
       "      <td>{'model__alpha': 0.0001, 'model__hidden_layer_...</td>\n",
       "      <td>0.878641</td>\n",
       "      <td>0.859223</td>\n",
       "      <td>0.856796</td>\n",
       "      <td>0.415049</td>\n",
       "      <td>0.881068</td>\n",
       "      <td>0.778155</td>\n",
       "      <td>0.181819</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>2.656595</td>\n",
       "      <td>0.501252</td>\n",
       "      <td>0.006267</td>\n",
       "      <td>0.000338</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>{'model__alpha': 0.0001, 'model__hidden_layer_...</td>\n",
       "      <td>0.621359</td>\n",
       "      <td>0.900485</td>\n",
       "      <td>0.912621</td>\n",
       "      <td>0.509709</td>\n",
       "      <td>0.565534</td>\n",
       "      <td>0.701942</td>\n",
       "      <td>0.170798</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.136293</td>\n",
       "      <td>0.465670</td>\n",
       "      <td>0.006000</td>\n",
       "      <td>0.000225</td>\n",
       "      <td>0.001</td>\n",
       "      <td>2</td>\n",
       "      <td>0.001</td>\n",
       "      <td>{'model__alpha': 0.001, 'model__hidden_layer_s...</td>\n",
       "      <td>0.837379</td>\n",
       "      <td>0.815534</td>\n",
       "      <td>0.876214</td>\n",
       "      <td>0.417476</td>\n",
       "      <td>0.398058</td>\n",
       "      <td>0.668932</td>\n",
       "      <td>0.214212</td>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>2.550768</td>\n",
       "      <td>0.360127</td>\n",
       "      <td>0.006207</td>\n",
       "      <td>0.000106</td>\n",
       "      <td>0.001</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>{'model__alpha': 0.001, 'model__hidden_layer_s...</td>\n",
       "      <td>0.752427</td>\n",
       "      <td>0.563107</td>\n",
       "      <td>0.864078</td>\n",
       "      <td>0.560680</td>\n",
       "      <td>0.526699</td>\n",
       "      <td>0.653398</td>\n",
       "      <td>0.131905</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>2.167006</td>\n",
       "      <td>0.939938</td>\n",
       "      <td>0.006290</td>\n",
       "      <td>0.000162</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>{'model__alpha': 0.0001, 'model__hidden_layer_...</td>\n",
       "      <td>0.825243</td>\n",
       "      <td>0.546117</td>\n",
       "      <td>0.351942</td>\n",
       "      <td>0.861650</td>\n",
       "      <td>0.453883</td>\n",
       "      <td>0.607767</td>\n",
       "      <td>0.202326</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>2.187955</td>\n",
       "      <td>0.705078</td>\n",
       "      <td>0.006425</td>\n",
       "      <td>0.000708</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>{'model__alpha': 0.001, 'model__hidden_layer_s...</td>\n",
       "      <td>0.803398</td>\n",
       "      <td>0.546117</td>\n",
       "      <td>0.347087</td>\n",
       "      <td>0.531553</td>\n",
       "      <td>0.776699</td>\n",
       "      <td>0.600971</td>\n",
       "      <td>0.169789</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.753331</td>\n",
       "      <td>0.952459</td>\n",
       "      <td>0.007793</td>\n",
       "      <td>0.001092</td>\n",
       "      <td>0.01</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>{'model__alpha': 0.01, 'model__hidden_layer_si...</td>\n",
       "      <td>0.427184</td>\n",
       "      <td>0.842233</td>\n",
       "      <td>0.847087</td>\n",
       "      <td>0.383495</td>\n",
       "      <td>0.410194</td>\n",
       "      <td>0.582039</td>\n",
       "      <td>0.214887</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2.097763</td>\n",
       "      <td>1.007147</td>\n",
       "      <td>0.007700</td>\n",
       "      <td>0.001959</td>\n",
       "      <td>0.01</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>{'model__alpha': 0.01, 'model__hidden_layer_si...</td>\n",
       "      <td>0.390777</td>\n",
       "      <td>0.550971</td>\n",
       "      <td>0.349515</td>\n",
       "      <td>0.526699</td>\n",
       "      <td>0.847087</td>\n",
       "      <td>0.533010</td>\n",
       "      <td>0.174876</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1.925540</td>\n",
       "      <td>0.819795</td>\n",
       "      <td>0.006040</td>\n",
       "      <td>0.000115</td>\n",
       "      <td>0.01</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>{'model__alpha': 0.01, 'model__hidden_layer_si...</td>\n",
       "      <td>0.157767</td>\n",
       "      <td>0.594660</td>\n",
       "      <td>0.587379</td>\n",
       "      <td>0.567961</td>\n",
       "      <td>0.521845</td>\n",
       "      <td>0.485922</td>\n",
       "      <td>0.166027</td>\n",
       "      <td>22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1.005035</td>\n",
       "      <td>0.571508</td>\n",
       "      <td>0.006140</td>\n",
       "      <td>0.000241</td>\n",
       "      <td>0.001</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>{'model__alpha': 0.001, 'model__hidden_layer_s...</td>\n",
       "      <td>0.417476</td>\n",
       "      <td>0.800971</td>\n",
       "      <td>0.385922</td>\n",
       "      <td>0.383495</td>\n",
       "      <td>0.378641</td>\n",
       "      <td>0.473301</td>\n",
       "      <td>0.164405</td>\n",
       "      <td>23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.952234</td>\n",
       "      <td>0.564018</td>\n",
       "      <td>0.006209</td>\n",
       "      <td>0.000414</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0005</td>\n",
       "      <td>{'model__alpha': 0.0001, 'model__hidden_layer_...</td>\n",
       "      <td>0.429612</td>\n",
       "      <td>0.793689</td>\n",
       "      <td>0.075243</td>\n",
       "      <td>0.398058</td>\n",
       "      <td>0.400485</td>\n",
       "      <td>0.419417</td>\n",
       "      <td>0.227796</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.367433</td>\n",
       "      <td>0.064478</td>\n",
       "      <td>0.007023</td>\n",
       "      <td>0.000460</td>\n",
       "      <td>0.01</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>{'model__alpha': 0.01, 'model__hidden_layer_si...</td>\n",
       "      <td>0.094660</td>\n",
       "      <td>0.060680</td>\n",
       "      <td>0.080097</td>\n",
       "      <td>0.089806</td>\n",
       "      <td>0.082524</td>\n",
       "      <td>0.081553</td>\n",
       "      <td>0.011650</td>\n",
       "      <td>25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.336036</td>\n",
       "      <td>0.063090</td>\n",
       "      <td>0.006105</td>\n",
       "      <td>0.000215</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>{'model__alpha': 0.0001, 'model__hidden_layer_...</td>\n",
       "      <td>0.094660</td>\n",
       "      <td>0.060680</td>\n",
       "      <td>0.075243</td>\n",
       "      <td>0.087379</td>\n",
       "      <td>0.082524</td>\n",
       "      <td>0.080097</td>\n",
       "      <td>0.011590</td>\n",
       "      <td>26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.324272</td>\n",
       "      <td>0.055400</td>\n",
       "      <td>0.006007</td>\n",
       "      <td>0.000260</td>\n",
       "      <td>0.001</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0001</td>\n",
       "      <td>{'model__alpha': 0.001, 'model__hidden_layer_s...</td>\n",
       "      <td>0.087379</td>\n",
       "      <td>0.063107</td>\n",
       "      <td>0.072816</td>\n",
       "      <td>0.087379</td>\n",
       "      <td>0.080097</td>\n",
       "      <td>0.078155</td>\n",
       "      <td>0.009262</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    mean_fit_time  std_fit_time  mean_score_time  std_score_time  \\\n",
       "24       0.970589      0.158883         0.006192        0.000285   \n",
       "6        1.043675      0.139572         0.006736        0.000917   \n",
       "15       0.918648      0.134335         0.006069        0.000223   \n",
       "16       1.322700      0.160295         0.006105        0.000126   \n",
       "7        1.277122      0.127310         0.006284        0.000298   \n",
       "25       1.227781      0.227956         0.006047        0.000276   \n",
       "12       1.112600      0.095687         0.006063        0.000236   \n",
       "21       1.118584      0.034868         0.006000        0.000193   \n",
       "13       1.810017      0.288030         0.006075        0.000220   \n",
       "3        1.252790      0.393719         0.006880        0.000538   \n",
       "4        1.724635      0.244262         0.006367        0.000545   \n",
       "22       1.442220      0.130967         0.006104        0.000220   \n",
       "0        1.733813      0.297156         0.007035        0.000360   \n",
       "18       1.496390      0.479643         0.006340        0.000201   \n",
       "26       2.656595      0.501252         0.006267        0.000338   \n",
       "9        1.136293      0.465670         0.006000        0.000225   \n",
       "17       2.550768      0.360127         0.006207        0.000106   \n",
       "23       2.167006      0.939938         0.006290        0.000162   \n",
       "14       2.187955      0.705078         0.006425        0.000708   \n",
       "1        1.753331      0.952459         0.007793        0.001092   \n",
       "5        2.097763      1.007147         0.007700        0.001959   \n",
       "8        1.925540      0.819795         0.006040        0.000115   \n",
       "10       1.005035      0.571508         0.006140        0.000241   \n",
       "19       0.952234      0.564018         0.006209        0.000414   \n",
       "2        0.367433      0.064478         0.007023        0.000460   \n",
       "20       0.336036      0.063090         0.006105        0.000215   \n",
       "11       0.324272      0.055400         0.006007        0.000260   \n",
       "\n",
       "   param_model__alpha param_model__hidden_layer_sizes  \\\n",
       "24             0.0001                               4   \n",
       "6                0.01                               4   \n",
       "15              0.001                               4   \n",
       "16              0.001                               4   \n",
       "7                0.01                               4   \n",
       "25             0.0001                               4   \n",
       "12              0.001                               3   \n",
       "21             0.0001                               3   \n",
       "13              0.001                               3   \n",
       "3                0.01                               3   \n",
       "4                0.01                               3   \n",
       "22             0.0001                               3   \n",
       "0                0.01                               2   \n",
       "18             0.0001                               2   \n",
       "26             0.0001                               4   \n",
       "9               0.001                               2   \n",
       "17              0.001                               4   \n",
       "23             0.0001                               3   \n",
       "14              0.001                               3   \n",
       "1                0.01                               2   \n",
       "5                0.01                               3   \n",
       "8                0.01                               4   \n",
       "10              0.001                               2   \n",
       "19             0.0001                               2   \n",
       "2                0.01                               2   \n",
       "20             0.0001                               2   \n",
       "11              0.001                               2   \n",
       "\n",
       "   param_model__learning_rate_init  \\\n",
       "24                           0.001   \n",
       "6                            0.001   \n",
       "15                           0.001   \n",
       "16                          0.0005   \n",
       "7                           0.0005   \n",
       "25                          0.0005   \n",
       "12                           0.001   \n",
       "21                           0.001   \n",
       "13                          0.0005   \n",
       "3                            0.001   \n",
       "4                           0.0005   \n",
       "22                          0.0005   \n",
       "0                            0.001   \n",
       "18                           0.001   \n",
       "26                          0.0001   \n",
       "9                            0.001   \n",
       "17                          0.0001   \n",
       "23                          0.0001   \n",
       "14                          0.0001   \n",
       "1                           0.0005   \n",
       "5                           0.0001   \n",
       "8                           0.0001   \n",
       "10                          0.0005   \n",
       "19                          0.0005   \n",
       "2                           0.0001   \n",
       "20                          0.0001   \n",
       "11                          0.0001   \n",
       "\n",
       "                                               params  split0_test_score  \\\n",
       "24  {'model__alpha': 0.0001, 'model__hidden_layer_...           0.953883   \n",
       "6   {'model__alpha': 0.01, 'model__hidden_layer_si...           0.956311   \n",
       "15  {'model__alpha': 0.001, 'model__hidden_layer_s...           0.941748   \n",
       "16  {'model__alpha': 0.001, 'model__hidden_layer_s...           0.949029   \n",
       "7   {'model__alpha': 0.01, 'model__hidden_layer_si...           0.949029   \n",
       "25  {'model__alpha': 0.0001, 'model__hidden_layer_...           0.898058   \n",
       "12  {'model__alpha': 0.001, 'model__hidden_layer_s...           0.919903   \n",
       "21  {'model__alpha': 0.0001, 'model__hidden_layer_...           0.922330   \n",
       "13  {'model__alpha': 0.001, 'model__hidden_layer_s...           0.922330   \n",
       "3   {'model__alpha': 0.01, 'model__hidden_layer_si...           0.946602   \n",
       "4   {'model__alpha': 0.01, 'model__hidden_layer_si...           0.917476   \n",
       "22  {'model__alpha': 0.0001, 'model__hidden_layer_...           0.895631   \n",
       "0   {'model__alpha': 0.01, 'model__hidden_layer_si...           0.873786   \n",
       "18  {'model__alpha': 0.0001, 'model__hidden_layer_...           0.878641   \n",
       "26  {'model__alpha': 0.0001, 'model__hidden_layer_...           0.621359   \n",
       "9   {'model__alpha': 0.001, 'model__hidden_layer_s...           0.837379   \n",
       "17  {'model__alpha': 0.001, 'model__hidden_layer_s...           0.752427   \n",
       "23  {'model__alpha': 0.0001, 'model__hidden_layer_...           0.825243   \n",
       "14  {'model__alpha': 0.001, 'model__hidden_layer_s...           0.803398   \n",
       "1   {'model__alpha': 0.01, 'model__hidden_layer_si...           0.427184   \n",
       "5   {'model__alpha': 0.01, 'model__hidden_layer_si...           0.390777   \n",
       "8   {'model__alpha': 0.01, 'model__hidden_layer_si...           0.157767   \n",
       "10  {'model__alpha': 0.001, 'model__hidden_layer_s...           0.417476   \n",
       "19  {'model__alpha': 0.0001, 'model__hidden_layer_...           0.429612   \n",
       "2   {'model__alpha': 0.01, 'model__hidden_layer_si...           0.094660   \n",
       "20  {'model__alpha': 0.0001, 'model__hidden_layer_...           0.094660   \n",
       "11  {'model__alpha': 0.001, 'model__hidden_layer_s...           0.087379   \n",
       "\n",
       "    split1_test_score  split2_test_score  split3_test_score  \\\n",
       "24           0.966019           0.951456           0.966019   \n",
       "6            0.966019           0.946602           0.961165   \n",
       "15           0.966019           0.946602           0.961165   \n",
       "16           0.963592           0.941748           0.951456   \n",
       "7            0.958738           0.932039           0.946602   \n",
       "25           0.958738           0.941748           0.958738   \n",
       "12           0.941748           0.944175           0.956311   \n",
       "21           0.932039           0.936893           0.953883   \n",
       "13           0.905340           0.946602           0.953883   \n",
       "3            0.915049           0.951456           0.885922   \n",
       "4            0.917476           0.927184           0.929612   \n",
       "22           0.890777           0.932039           0.927184   \n",
       "0            0.837379           0.837379           0.854369   \n",
       "18           0.859223           0.856796           0.415049   \n",
       "26           0.900485           0.912621           0.509709   \n",
       "9            0.815534           0.876214           0.417476   \n",
       "17           0.563107           0.864078           0.560680   \n",
       "23           0.546117           0.351942           0.861650   \n",
       "14           0.546117           0.347087           0.531553   \n",
       "1            0.842233           0.847087           0.383495   \n",
       "5            0.550971           0.349515           0.526699   \n",
       "8            0.594660           0.587379           0.567961   \n",
       "10           0.800971           0.385922           0.383495   \n",
       "19           0.793689           0.075243           0.398058   \n",
       "2            0.060680           0.080097           0.089806   \n",
       "20           0.060680           0.075243           0.087379   \n",
       "11           0.063107           0.072816           0.087379   \n",
       "\n",
       "    split4_test_score  mean_test_score  std_test_score  rank_test_score  \n",
       "24           0.973301         0.962136        0.008209                1  \n",
       "6            0.978155         0.961650        0.010457                2  \n",
       "15           0.978155         0.958738        0.013205                3  \n",
       "16           0.956311         0.952427        0.007298                4  \n",
       "7            0.970874         0.951456        0.012935                5  \n",
       "25           0.966019         0.944660        0.024628                6  \n",
       "12           0.961165         0.944660        0.014351                6  \n",
       "21           0.953883         0.939806        0.012414                8  \n",
       "13           0.939320         0.933495        0.017543                9  \n",
       "3            0.963592         0.932524        0.028281               10  \n",
       "4            0.958738         0.930097        0.015150               11  \n",
       "22           0.934466         0.916019        0.018838               12  \n",
       "0            0.895631         0.859709        0.022425               13  \n",
       "18           0.881068         0.778155        0.181819               14  \n",
       "26           0.565534         0.701942        0.170798               15  \n",
       "9            0.398058         0.668932        0.214212               16  \n",
       "17           0.526699         0.653398        0.131905               17  \n",
       "23           0.453883         0.607767        0.202326               18  \n",
       "14           0.776699         0.600971        0.169789               19  \n",
       "1            0.410194         0.582039        0.214887               20  \n",
       "5            0.847087         0.533010        0.174876               21  \n",
       "8            0.521845         0.485922        0.166027               22  \n",
       "10           0.378641         0.473301        0.164405               23  \n",
       "19           0.400485         0.419417        0.227796               24  \n",
       "2            0.082524         0.081553        0.011650               25  \n",
       "20           0.082524         0.080097        0.011590               26  \n",
       "11           0.080097         0.078155        0.009262               27  "
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results.sort_values(\"rank_test_score\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         dws       0.88      0.94      0.91       236\n",
      "         jog       0.98      0.94      0.96       265\n",
      "         sit       0.99      1.00      0.99       663\n",
      "         std       1.00      1.00      1.00       615\n",
      "         ups       0.93      0.92      0.92       298\n",
      "         wlk       0.98      0.97      0.98       666\n",
      "\n",
      "    accuracy                           0.97      2743\n",
      "   macro avg       0.96      0.96      0.96      2743\n",
      "weighted avg       0.97      0.97      0.97      2743\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_train, grid_search.predict(X_train), target_names=label_encoder.classes_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x145912ed0>"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(grid_search, X_train, y_train,\n",
    "                      display_labels=label_encoder.classes_,\n",
    "                      cmap=plt.cm.Blues,\n",
    "                      normalize=\"true\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "pycharm": {
     "is_executing": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         dws       0.85      0.93      0.89       123\n",
      "         jog       0.98      0.92      0.95       119\n",
      "         sit       0.99      0.99      0.99       341\n",
      "         std       0.99      1.00      0.99       285\n",
      "         ups       0.95      0.90      0.92       138\n",
      "         wlk       0.97      0.97      0.97       346\n",
      "\n",
      "    accuracy                           0.97      1352\n",
      "   macro avg       0.96      0.95      0.95      1352\n",
      "weighted avg       0.97      0.97      0.97      1352\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, grid_search.predict(X_test), target_names=label_encoder.classes_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x144c0f1d0>"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUMAAAEKCAYAAACIZDejAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdeXxMV//A8c83iyUhC0lIEFutpTS0VVtRlFJaqtXN0qdPF1Xd6K/VVlHdqKp6dNH1KR5K0WpRWxG1S+xrUbEkJEFil2Tm/P6YEZmQZCSZyeD77mtenTv3zDnfuff65tztXDHGoJRSNzqvog5AKaU8gSZDpZRCk6FSSgGaDJVSCtBkqJRSgCZDpZQCNBkqpa4xIvKdiCSKyNYc5ouIfCYie0Rks4hEOVOvJkOl1LXmB6BDLvM7AjXsr6eBL5ypVJOhUuqaYoyJBo7nUqQr8KOxWQ0EiUh4XvX6FFaARUWK+RspEVzUYWRqWKN8UYfg8aSoA8jG6oE3YYkHLaQDcftJTk4uUETeAZWNyTjnVFlzLmkbcD7LRxOMMROuorkKwMEs04fsnyXk9qVrPxmWCKb4bc8XdRiZouf+X1GHcBkvT/qXBXh5eVY8F9ItRR3CZbw9aBm1bHp7geswGecoXushp8qe3zj+vDGmcYEbvUrXfDJUSl0LBMRtR+UOA5WyTFe0f5YrPWaolHI9Aby8nXsV3Gygl/2schMg1RiT6y4yaM9QKeUuhXS4RkSmAK2AEBE5BLwD+AIYY74E5gL3AnuAs0BfZ+rVZKiUcoPC2002xjySx3wDXPWJBE2GSin38LATedlpMlRKuZ7gzhMo+aLJUCnlBqI9Q6WUAgrrTLHLaDJUSrmBW68zzBdNhkop1xN0N1kppQDtGSqllO4me4i7G1fjg2fb4u3txcR5G/l02mqH+ZXCAhj3SidCAv04ceocz4z8jfjkU1QKC2DikO54eQk+Pl58/WsM38/ZkK8YFq/azptjZmKxWnm8y5282Kudw/wLaek8P2wSm3YdpEyAP1+P6ENkRFmOp57hyTe+ZcOOA/TsdAcfDeyR+Z209Axe//hnVsT+jZeXMPiZztzXpmGuMbzxyQys9hhe6t3+shj6DZvIpp0HCQ7059sRfYmMKAvAmB8WMPm3VXh5efHhqw/SpkkdAL6Y8icTf12FiFC3ejjj3n6cEsV9+Xr6Mr6aupR/DiWze/4HlA0qla/ldiWLVm7njdE/Y7FaeaJrU17u0z7vL+XDn6t38PanM7FYrDx2XxNeuGydZfDCu5PYbF9eX73bm8jwssRuj2PQRz8BYIxh4L86cO9dDTh/IZ37+31GWnoGGRYrnVs34LWn7nU6Hk/YhvJNAG89gQKAiAwFThtjPnZXm2AbIWXU8+154I2pxCef5M9xfZi3+m92HTiWWWb4v9swddFWpi7aQosGlRnStxXPjvqNI8dP0/7lH0lLt+BfwpeVXz3FvFV/c+T46auKwWKx8vrH05n+2fNEhAXRvu/HdGhRj1pVLw2xNnn2aoIC/Fj38xBmLYxh+PjZfPNeX4oX8+H1pzuxc18CO/Y53l455ocFhASXYs30t7FarZw4eTbXGF4bNZ0Z42wxtO0zig4t6lO72qUYJs1eRVBpP9bPeIeZC2IYNv5Xvn3vSXbuS2DWwhhWTBnMkeRUuvUfz9rpb3P02Ekm/LSMlVPfpGSJYjw5+DtmLozh0c5NuOOWatzTrB5d+n12VcvKmWU5aOQ0Zv2nPxHlgmjTexQdWzr+jsJq542PpzNtbD/Cw4Lo8K/RtG9Rn1pVLw3R9r/fVhFUuiSrp7/NLwtjGfH5b0x4tw+1q4Uz/9tX8fHx5mhyKm16jaR9s3oUL+bDjHH98fcrTnqGhS7PjuXuJnVpVK+KU/EU9TZUYB5+zNCz+62FoFGtCPbFnyDuSArpGVZmLt3BvXfWdChTq3IIyzftB2D5pjg63lkDgPQMK2n24Z2K+frke+ip2O1xVKkYSpUKIRTz9eH+dlHMi97iUGbe8i08fK9tqKT7Wjdk+frdGGPwL1mcJg2rU7yY72X1/u+31bzY29Y78PLyyrX3Fbs9jqoVQzJjeKBdo8tjiN5Cz053ANClTUOi19limBe9hQfaNaJ4MV8qR4RQtWIIsdvjAMiwWDl/IZ2MDAvnzqcRHhIIwC21KmX2KgtTzLb9VKsUQpWKtt/RrV0Uc5dtLvR2NmyPo2rFUCpfXGdto5i/3HF5zV++lYc62tZZ59YN+Mu+zvxKFMPHx9YLOp+WkZkDRAR/v+IApGdYyMiwOJ0fPGEbKhj7brIzryLi0pZF5E0R2S0ifwG1AC8RibHPayAiRkQi7dN7RcRPRHqIyFYR2SQi0QWNIbxsKQ4nncycjk8+RXhIaYcy2/Yl0rlZLQA6N6tJgH9xgkuXBKBCaGn++uJfbJ30PGOnrb7qXiFAQlIKFcKCMqcjwoJISEp1KHMkKZUK5WxlfHy8CShVguOpZ3KsM/WU7S/4h1/NoU2vkTw5+DsSj53MsXxCYgoVyl0aBNcWQ0q2OFOJCMsaQ0mOp56xxZ/9u4kpRIQF0f+xu2nQdQh1O71FQKmStLbvPrtKQlKqYyzlgi9bloXVTkS5S+ssPPTydZaQlEKEPRYfH29K+19aZ7Hb9tPysQ9o/cSHjHztoczkaLFYubv3SOp1epOWt9Ui6uYqTsZT9NtQgYk49yoiLkuGItII6Ak0xDaCxG2AFSghIgFAC2A90EJEKgOJxpizwBDgHmNMA6BLDnU/LSLrRWS9Sct5ZTvr7Ql/0qx+JMvG96VZ/UgOJ53EYrUCcDjpFM2f+5ZGfb+kZ7v6hAb5Fbi9wpBhsRKfmMJtt1Tlzx9f47Z6VRg67he3xpBy8ixzozcTO2so2+aM4My5C0ybt86tMXiqqJurED35Df749lU++3ER5y+kA+Dt7cXi/77Ghl+GsWFHHDv2xhdZjG7fhm7gnmELYJYx5qwx5iS2McYAVgLNgJbA+/b/twCW2+evAH4QkX8DVzziaoyZYIxpbIxpLMX8cw0i4dhpKoQGZE5HhJQmIfmUQ5kjx0/T692Z3PX894z4YRkAJ89cuKzMjv1J3FmvElcrPDSIw4mXemHxiSmEhwY6lCkfGsjho7YyGRkWTp4+T5nAnH9bmUB//EoUo3OrBgB0uftWNu86lHMMYUEcPnoiWwxBjmVCA4lPzBrDOcoE+tviz/7dsCCWrdtF5YiyhASXxtfHm86tG7B2y768FkeBhIcGOsZy9MRly7Kw2ok/emmdJSRdvs7CQ4OIt8eSkWHh1JnL11nNKuXxL1mcndmO1QWW9qNZVA2WrNnpZDxFvw0ViLO9wuuxZ5iLaGzJrzLwK9AAaI49GRpjngXewjZSbYyIFOjAU+yueKpXCCayXCC+Pl50a1WHeav/dihTJqBk5jp4ueedTF5gOwYVEVKaEsVs55gCS5Wgyc2V2HMot+fQXNmtdSL552AScfHHSEvP4JeFsXRoUd+hTIcW9fhp7loAfluykeaNayC5bBgiQvvm9VgRuweA6HW7qVk15+ev3Fonkn0Hk4iLTyYtPYNZC2Po2DJ7DPWZOmcNALP/3EiLxjURETq2rM+shTFcSEsnLj6ZfQeTiKpbmQrlglm/dT9nz6dhjLHFUMW1z4CJqluZvQeSiDts+x0zF8bSseUthd5OwzqR7DuUZZ0tiqV983oOZdq3qMe0ebZ19vuSTTRrZFtncfHHyMiwHWs+mHCcPQeOUim8DMknTmfump67kEb0ul3cVDnMqXg8YRsqMPcN7povYhv6ywUV255V+gNwB7az1rHAV8DP2BJitDHmcRGZC9QDGhhjTohIdWPMXnsd64B/G2M25tSOV0BFk9czUNrdVp33n22Lt5cwecFmRk9ZyRu9WrBxdwLzVu+hS/NaDHmyFcbAyi0HGDR+AWnpFlpFVWHEv+/GYBCEr2fH8N95OYYCQFIOz0BZuHIbb42ZidVq5ZHOTXil7z18OGEODWtH0qFlfc5fsF3WsmX3IYID/Jjwbh+qVAgBIOr+oZw6e5609AwCS5Vk+mf9qFU1nIMJx+k3bCInT52jbHApPnvrUSqWL3P5MrL/g1i4YhtvjpmBxWp49L4mvNr3Hj74ag4N60TS0R7Dc0N/ZMvuQwQF+PHNiL6ZMYz+fj7/+2013t5evP9yN9o2vRmADyfMYdaiWHy8valfsyJj33yE4sV8+eqnpYybuJjE4ycJCS5Fu6Y3M/bNR23xFPD5HgtWbGPwJz9jsRge69KEgU/m9tTIvOX0DJRFK7cxZOwsLBbbOnupT3s++nouDWtX4p4WtuXVf/gkttqX11fDe1O5QgjT561j3KRF+Pp44yXCK33voeNdt7B9z2EGvDsZi9WK1WrocvetvJpD7Fd6BkpRbUMtm95ObMz6Aq00r8BIU7zZq06VPT/vpZiieAaKy5Ih2E6gAL2BROAAEGuM+VhEDgLvGmMmiMhgoKcx5hb7d2Zie96pAIuBl0wuQTqTDN0pp2RYlPSBULnTB0LlrtCSYfOBTpU9P/fFIkmGLr3O0BjzHvDeFT6vlOX9+9iOHV6c7ubKmJRSRUDHM1RKKdDb8ZRS6iIdz1AppfD42/E0GSqlXE90N1kppWy0Z6iUUuR6Abgn0GSolHI526j/mgyVUjc6EcSDLiS/Ek2GSim30J6hUkqhyVAppQBNhkopZT+DUtRB5O6aT4a33FSOP2cPKuowMoXe90lRh3CZxNmvFHUIDjxt1Jrivp53m5jV6rrRpIqCINozVEopsD1wypNpMlRKuYX2DJVSSo8ZKqWUjaf3DD17J14pdV24eALFmZdT9Yl0EJFdIrJHRF6/wvxIEVkiIhtEZLOI3JtXnZoMlVJuIV7i1CvPekS8gfFAR6Au8IiI1M1W7C1gmjHmVmzPb/88r3o1GSqlXE8ozJ7h7cAeY8w+Y0waMBXomq2MAS4+MD0QiM+rUj1mqJRyi6s4ZhgiIuuzTE8wxkzIMl0BOJhl+hC2RxJnNRRYICIvAP5A27wa1WSolHKLq0iGyYXwqNBHgB+MMaNF5E5goojUM8ZYc/qCJkOllMsV8h0oh4FKWaYr2j/L6l9ABwBjzCoRKQGEYHuG+xXpMUOllHuIk6+8rQNqiEhVESmG7QTJ7GxlDgB3A4hIHaAEkJRbpdozVEq5nhTe7XjGmAwR6Q/MB7yB74wx20RkOLDeGDMbeBX4WkRexnYypY8xJtcbvjUZKqXcojAvujbGzAXmZvtsSJb324FmV1OnJkOllHt49g0oN0YyXLJ6B0PGzsRqNTzSuQn9n3A8y34hLYMXR0xiy65DBAf48cXw3lQKL8vBhGO0euxDqkWGAhB1cxU+GvRQgeO5u1EVPnimDd5ewsT5W/h0+lqH+ZXCAhj30j2EBPpx4tR5nhk1h/hjp6lXLZTRz7ejtF8xrFbD6J9WMyt6V75i+HPVdt78dCYWi5XHu9zJgF7tHOZfSEun//BJbNp5kDKB/kwY0YfI8LIsXbuTEZ/PJj3dgq+vN+/0v58WjWsCcH+/zzh67CQlivsCMO3TfoSWKZ2v+HKzaOV23hj9MxarlSe6NuXlPu0LvQ1PiWnxqu288ckMrFbbenqpt2O9F9LS6TdsIpt2HiQ40J9vR/QlMqIsAGN+WMDk31bh5eXFh68+SJsmdQD4YsqfTPx1FSJC3erhjHv78cx1BvD66J/532+rOLB0dKH8hos8/XY8tydDEVlpjGnqrvYsFitvfvIzU8Y8R3hYEPc+9Qntm9ejZtXymWWm/L6awNJ+rPjpLX5dFMt7X/zGl8P7AFC5QlkW/vBaocXj5SWM6teWB96cTnzyKf789HHmrd7LroPHMssM/9ddTF28namLt9GiQSWG9G3Bsx/P49yFDJ4bPZd98SmUL+PPks+eYHHMfk6euXBVMVgsVv5v9HSmj32eiLAg2j/5Mfe0qEetquGZZSb/Zlsma38ewqyFMbw7fjZfj+hL2UB/Jo16hvKhgezYG8/DL33B5t/ezfzeF0N70bBOZMEXVC6xDxo5jVn/6U9EuSDa9B5Fx5b1qV0tPO8vX2MxWSxWXhs1nRnjbOupbZ9RdGjhWO+k2asIKu3H+hnvMHNBDMPG/8q37z3Jzn0JzFoYw4opgzmSnEq3/uNZO/1tjh47yYSflrFy6puULFGMJwd/x8yFMTzauQkAG3YcIOXk2QLFfSVXc6tdUXH72WR3JkKADTviqFIxhMoVQijm60PXtrcy/68tDmUW/LWFHh1vA6BTqwb8FfM3eRxrzbdGNcuzL/4EcUdSSc+wMjN6J/feWd2hTK3IsizfdACA5ZsO0rHJTQDsPXyCffEpABw5fobklLOEBJa86hhit8dRtWIoVezL5IG2UfwR7bhM/li+hYfvvR2A+1o3ZPn63RhjqF+rEuVDAwGoXS2c8xfSuZCWftUx5FfMtv1UqxRClYq22Lu1i2Luss1ua9+dMdnWU8il9dSuEfOyrad50Vvo2cl2vXGXNg2JXmdbT/Oit/BAu0YUL+ZL5YgQqlYMIXZ7HAAZFivnL6STkWHh3Pk0wkNs69NisTL0s18Y+kL2mzkKR2Hem+wKbk+GInJabEaJyFYR2SIiD9vneYnI5yKyU0QWishcEXmwIO0dSUolIiw4czo8NIgjSak5lvHx8SbAvwQnUs8AcCDhOO37jqJ7/3Gs2bS3IKHY2i9bmsPJpzKn45NPE17WcVdy2z9JdG5WA4DOTWsQ4Fec4NIlHMpE1SyPr483/ySkXHUMR5JSqBAWdCmmsCASrrBMKpSzlfHx8aZ0qRIcty+Ti35fspH6tSpSvNilXawXR0ymda+PGP3dHy75g5KQlEqFcpfWZ0S54MtidzdXxZSQmOJYb1gQCUmO6zshKZWIsEvrKaBUSY6nniEh6QrfTUwhIiyI/o/dTYOuQ6jb6S0CSpWktX33+Zvp0XRoWY/y9uRY2Arr3mRXKapjht2AhkADbBdCrhORaGxnf6pgu/k6DNgBfJf9yyLyNPA0QMVKrtslCysbyNoZ71Am0J/NOw/y5OBvWTLxdUr7l8j7ywXw9jdLGfnc3Tza9mZWbj3E4eRTWLIMA18u2J8vB95Lv9HzcFEHNk879yUw/PPZTPu0X+ZnXwztRXhYEKfPnKfv4G+ZNm9dZu9SeYaUk2eZG72Z2FlDCSztR983bOupReOa/Lp4A7O/GOCytnU3+cqaA1OMMRZjzFFgGXCb/fPpxhirMeYIsORKXzbGTDDGNDbGNC4bEpJrQ+VDA4lPPJE5nZCUkrmbd6UyGRkWTp45T3CgP8WL+VAm0B+AW2pXokpEWfYdzPECdqckHDtFhZBLPcGIkFIkHDvlUObI8TP0em82d70wkRH//Qsg87hg6ZLF+GlYN0b89y/W70rIVwzlQ4M4nHiph5GQmEL4FZbJ4aO2MhkZFk6dPp+5LOITT9Dn9W/4z9tPULViaOZ3wu09lFL+JejevjEb7LtlhSk8NJDDRy+tz/ijJy6L3d1cFVN4WJBjvYkphIcGOZYJDSQ+8dJ6Onn6HGUC/QkPvcJ3w4JYtm4XlSPKEhJcGl8fbzq3bsDaLfvYsusg/xxKovGDw2l4/zucPZ9O4+7DCvwbMhXuQA0ucd3fgdKwdiT/HEzmQPwx0tIz+HXRBto3q+dQpn2zekyftw6AOUs30SyqBiLCsROnsVhstzLGHU7mn0PJmWfq8it29xGqRwQTWS4QXx8vurWszbzVjrvfZQJKcnGbePmhO5i8YCsAvj5eTHy7K1MXb2P2it35juHWOpHsO5hEnH2ZzFoUyz0t6juUuad5PX6aazvL/duSjTRvZFsmqafO8uirX/FWvy7c0aBaZvmMDAvHUk4DkJ5hYcGKrS45qRFVtzJ7DyQRdziZtPQMZi6MpWPLWwq9HU+I6dJ6stU7a2EMHVs6rqcOLeozdc4aAGb/uZEWjWsiInRsWZ9ZC2O4kJZOXHwy+w4mEVW3MhXKBbN+637Onk/DGEP0ut3UrFKe9s3rsWPe+2z8ZRgbfxmGXwlf1s94p8C/4SIBRJx7FZWi2k1eDjwjIv8FygAtgUFAcaC3/fNQoBXwv4I05OPjzYhXuvPoK19itVp5uNMd1KoWzqhv5tKgdiTtm9ejZ+cmDHh3Es0eHkFQgB+fD+0FwOpNe/n4m3n4+Hjh5eXFBwN7EBzgX5BwsFgNr32xmBkjuuPt5cXkBVvYeeAYbzzejI1/H2Hemr00r1+JIX1aYDCs3HqIQeMXA/BAi1o0rVeRMqVL8mhbW0LvN2YeW/flepfRFZfJh68+yMMvfY7FauXRzk2oXS2cDyfMoWGdSDq0qM9j993J88MmcvuDwwkO8OOrd/sA8O3Py9l/KJnR3/3B6O/+AGyX0PiVLMbDL31OeoYVq9VKy9tq8UTXwj9X5uPjzcjXHqL7gPFYLIbHujShTvWiO5Psyph8fLz5aGAPegz4HIvV8Oh9tvX0wVe29dSxZX0e73Inzw39kcbdhxEU4Mc3I/oCtpNbXdtG0bTn+3h7ezFyUA+8vb1oXK8KXdo0pHWvj/Dx9qZ+zYr0vt8d5zQ9/2yyuOqsaY4NipzCNs7YSGyDMxpghDHmJxHxwjYIYytsQ/QI8JExZmFO9TWMamT+XL7G5XE7q8IDnxZ1CJfxtEeF+vpc9zskBeZJjwptfudtxMasL1AmK1G+pqnce5xTZXeP7BBTCKPWXDW39gxFpCxw3H6P4CD7K5MxxioiA40xp+1l1wJbrlCVUupaUsS7wM5wWzIUkQhgKfBxHkV/F5EgoBjwrv1EilLqGibYbjjwZG5LhsaYeKCmE+VauT4apZS7ac9QKaXw/OsMNRkqpVxPjxkqpZRt2P/CGtzVVTQZKqXcQnuGSimFHjNUSik9ZqiUUnDx3mTPzoaaDJVSbuHhuVCToVLKPfQOFKWUEt1NdjlvEfyKe87PODF3YFGHcJngO14s6hAcHFvlWSP7eGKPxZNiKoxILo5n6Mk8J4sopa5jnj+eoSZDpZRbeHgu1GSolHID8axd/yvRZKiUcjm9zlAppew0GSqlFHrMUCmlAO0ZKqWUDtSglFJwcXBXz86GmgyVUm7h5eFdQ88eh1spdd0Qce7lXF3SQUR2icgeEXk9hzIPich2EdkmIv/Lq07tGSqlXE4KcaAGEfEGxgPtgEPAOhGZbYzZnqVMDeANoJkx5oSIhOVVb47JUEQCcvuiMeaks8ErpVQhHjK8HdhjjNkHICJTga7A9ixl/g2MN8acADDGJOZVaW49w22AwXHQiovTBoi8mug92aKV23lj9M9YrFae6NqUl/u0v6HiubtJbT54pRveXl5MnL2aT39c5DC/Uvlgxr31KCFBpThx8gzPDJ1IfGIqAEOfv4/2zW4GYNR385m1aIPT7S5etZ03PpmB1Wrl8S538lJvx995IS2dfsMmsmnnQYID/fl2RF8iI8oCMOaHBUz+bRVeXl58+OqDtGlSB4AvpvzJxF9XISLUrR7OuLcfp0RxX5at28XQcb9gtRr8SxbnP0Mep1ql0Hwvs+xu9G3IGVdxAiVERNZnmZ5gjJmQZboCcDDL9CHgjmx11AQQkRWANzDUGPNHrvHlNMMYU8kYE2n/f6Vs0y5JhCLyjYjUtb8f7Io2srNYrAwaOY3pY/uxetpbzFgQw859Ce5o2iPi8fISRg3qQY+XvqJJzw/o3j6KWlXLOZQZPqArU+eupfnjHzHy2/kM6XcfAO2b1eWWWpVo8cRI2j75Cf0fa0Np/+JOtWuxWHlt1HSmffocK6e+ycwr/M5Js1cRVNqP9TPe4bmerRk2/lcAdu5LYNbCGFZMGcz0sc8xaOQ0LBYr8YkpTPhpGYt/GMSKKYOxWA0zF8YAMOijn/hyWG+WTXqd7vc0YvT3uf67uCo3+jbkDMF2RtmZ/4BkY0zjLK8JeVR/JT5ADaAV8AjwtYgE5fYFp06giEjPi8lJRCqKSKN8BJcnY8xTWfb73ZIMY7btp1qlEKpUDKGYrw/d2kUxd9lmdzTtEfE0qluZfYeSiIs/RnqGhZkLY7m3ZX2HMrWqlmf5+r8BWB7zNx3t82tVLc/KjXuwWKycPZ/Gtj3x3G3voeUldnscVSuGUKWC7Xc+0K4R86K3OJSZF72Fnp1sf/C7tGlI9LrdGGOYF72FB9o1ongxXypHhFC1Ygix2+MAyLBYOX8hnYwMC+fOpxEeEmirTIRTZ84DcPL0ecpf/LwQ3OjbkLO8xLmXEw4DlbJMV7R/ltUhYLYxJt0Y8w+wG1tyzDm+vFoVkf8ArYEn7B+dBb50KuTc6/UXkTkisklEtorIwyKyVEQai8iHQEkR2SgikwvaVm4SklKpUC44czqiXDAJSamubNKj4gkPC+Tw0ZTM6fjEFMJDHRPFtr/j6dy6AQCdW91CgH8JggP82Pr3Ydo2qUPJ4r6UCfSnRaObHGLPTUJiiuPvDAsiISnFsUxSKhFhtj/mPj7eBJQqyfHUMyQkXeG7iSlEhAXR/7G7adB1CHU7vUVAqZK0tifnsYMfoefLX1Cv89tMm7eOF3u1u4qllMdvucG3IaeIbTxDZ15OWAfUEJGqIlIM6AnMzlbmF2y9QkQkBNtu877cKnWmZ9jUGPMMcB7AGHMcKOZMxHnoAMQbYxoYY+oBmfstxpjXgXPGmIbGmMeyf1FEnhaR9SKyPik5qRBCUbl5+7NfaHZrdZb9OIhmUTdxODEFi9WwZM0uFq7czvxvXuKbd3uzbst+LFZrkcWZcvIsc6M3EztrKNvmjODMuQtMm7cOgC+mLmHqmOfY+vu7PNr5Dt4eO6vI4rxRFdalNcaYDKA/MB/YAUwzxmwTkeEi0sVebD5wTES2A0uAQcaYY7nV68ylNeki4oXtpAkiUhYojC1+CzBaRD4CfjfGLHf21Lv9GMIEgEaNGpuCBBEeGsjhoycyp+OPnrisZ+RO7o4nITGVCuUuHUqx9dAcexFHkk/S6/XvAPAvWYz7Wjfg5OlzAIz+YSGjf1gIwNfDe7H3gHN/nMLDghx/Z2IK4aGOh3TCQwOJt/cgMzIsnDx9jjKB/oSHXuG7YUEsW7eLyhFlCQkuDUDn1g1Yu2UfbZrUZtvf8f7AQEcAACAASURBVDSuVwWAB9pF0ePFL5yK06nfcoNvQ84QCveia2PMXGButs+GZHlvgFfsL6c40zMcD8wAQkVkGPAX8JGzDeTEGLMbiMKWFEeIyJA8vuISUXUrs/dAEnGHk0lLz2Dmwlg6trylKEIpknhidxygeqVQIsPL4OvjTbd2UcyL3upQpkygf+buy8u92zH5t9WA7eRLcIAfADffFMHNN0Xw55qdTrV7a51I9h1MIi7e9jtnLYzJPBZ5UYcW9Zk6Zw0As//cSIvGNREROrasz6yFMVxISycuPpl9B5OIqluZCuWCWb91P2fPp2GMIXrdbmpWKU9QaT9Onj7HngO2qyuWrt1FzSrlLospv270bchZXl7i1Kuo5NkzNMb8KCIxQFv7Rz2MMVtz+44zRCQCOG6MmSQiKcBT2Yqki4ivMSa9oG3lxsfHm5GvPUT3AeOxWAyPdWlCnerhrmzSo+KxWKy89vEMZnz2HN5eXkz+bTU7/znCG093ZOOOg8xbvpXmjW5iSL/7MMawcsNeBo2aDoCvjzdzJ9geNnXqzHmefmciFotzOw0+Pt58NLAHPQZ8jsVqePS+JtSuFs4HX82hYZ1IOrasz+Nd7uS5oT/SuPswggL8+GZEXwBqVwuna9somvZ8H29vL0YO6oG3txeN61WhS5uGtO71ET7e3tSvWZHe9zfFx8ebMYMfoc/r3+AlQlCAH5+9ddnRl3y70bchZ1zN3SVFRWy9yTwKidwCNMe2q7zCGFPgU1Micg8wCtsudzrwHPAxMNAYs96++9wFiL3SccOLGjVqbFasWZ/TbIU+HS8vnj6AQFFrdkdjYmLWF2ghlala17Qb6ty50Gl9omKMMY0L0l5+5NkzFJE3gUeBWdh2/f8nIpONMR8UpGFjzHxsBzmzapVl/v8B/1eQNpRSnsPT/+Q4cwKlF3CrMeYsgIi8B2wACpQMlVI3luthcNeEbOV87J8ppZRTbGeTizqK3OU2UMMYbMcIjwPbRGS+fbo9toselVLKOXJtD+568YzxNmBOls9Xuy4cpdT16prdTTbGfOvOQJRS169rejf5IhGpDrwH1AVKXPzcGFPThXEppa4znt4zdOYOlB+A77El947ANOAnF8aklLoOiZOvouJMMvSzXxOIMWavMeYtbElRKaWcIgLeXuLUq6g4c2nNBftADXtF5Fls44aVdm1YSqnrjafvJjuTDF8G/IEB2I4dBgJPujIopdT1x8NzoVMDNayxvz3FpQFelVLKaYJ4/HOTc7voehb2MQyvxBjTzSURKaWuP9fAqDW59Qz/47YolEudWDO2qENwEHxb/6IOwcGJdbqpu8M1e8zQGLPYnYEopa5fAnhfq8lQKaUK0zV/B4pSShWG6yYZikhxY8wFVwajlLo+2Yb99+xs6Mxzk28XkS3A3/bpBiIyzuWRKaWuK4X4EHnXxOdEmc+AzsAxAGPMJmwPlVdKKacV1nOTXcWZ3WQvY0xcti6uxUXxKKWuQwL4ePhusjPJ8KCI3A4YEfEGXgB2uzYspdT1xsNzoVPJ8Dlsu8qRwFFgkf0zpZRyisg1fDveRcaYRKCnG2JRSl3HPDwXOjXS9ddc4R5lY8zTLolIKXVduh6uM1yU5X0J4AHgoGvCUUpdjwSKdOBWZzizm+wwxL+ITAT+cllESqnrTxFfQ+gMZ64zzK4qUK6wAylKi1Zu57buw4l6YChjflhQ1OFoPHkY9/Zj7J7/ASunDi7qUDJ52jLytHjANqahM/8VFWfuQDkhIsftrxRgIfBGYTQuIi+JiF8O8/qIiMvHVrJYrAwaOY3pY/uxetpbzFgQw859Ca5uVuMpgCm/r+bBAeOLNIasPG0ZeVo8cOlRodfsHShiu9K6ARBqfwUbY6oZY6YVUvsvAVdMhu4Ss20/1SqFUKViCMV8fejWLoq5yzZrPB4aD8DKDXs5cfJskcaQlactI0+L56JrOhkaYwww1xhjsb9yHPk6LyLiLyJzRGSTiGwVkXeACGCJiCyxl+krIrtFZC3QLL9tXY2EpFQqlAvOnI4oF0xCUqo7mtZ4rhOetow8LZ6LRMSpV1Fx5mzyRhG51RizoYBtdQDijTGdAEQkEOgLtDbGJItIODAMaASkAkuAK7YpIk8DTwNUiowsYFhKKVezPSq0qKPIXY7hicjFRHkrsE5EdolIrIhsEJHYfLS1BWgnIh+JSAtjTPY/VXcAS40xScaYNHJ5UL0xZoIxprExpnFoSGg+QrkkPDSQw0dPZE7HHz1BeGhggerUeG4snraMPC2ei7zsd6Hk9XKGiHSw56Q9IvJ6LuW6i4gRkcZ5xpfLvLX2/3cBagH3Aj2AB+3/vyrGmN1AFLakOEJEhlxtHa4QVbcyew8kEXc4mbT0DGYujKVjy1s0Hg+NxxN52jLytHigcE+g2MdIGA90BOoCj4hI3SuUKw28CKzJPu9KcttNFgBjzF5nKsqLiEQAx40xk+xnpZ/C9vjR0kAytoDHikhZ4CS2hLupMNrOjY+PNyNfe4juA8ZjsRge69KEOtXDXd2sxlMA34zoQ7NGNSgbVIqtv7/LhxPmMmn2qiKLx9OWkafFc1EhHg68HdhjjNlnq1emAl2B7dnKvQt8BAxyptLckmGoiLyS00xjzCfONJBFfWCUiFiBdGyDPdwJ/CEi8caY1iIyFFgFpAAbr7L+fGvf7GbaN7vZXc3lSePJ3VNv/VDUIVzG05aRp8UDgpfz1xCGiMj6LNMTjDETskxXwPEuuEPYDrNdak0kCqhkjJkjIgVOht5AKSicqyCNMfOB+dk+Xg+My1Lme+D7wmhPKeU5hKvqGSYbY/I8xpdjWyJewCdAn6v5Xm7JMMEYMzy/ASmlVCYBn8K7iPAwUCnLdEX7ZxeVBuoBS+2X6pQHZotIF2NM1h6ngzyPGSqlVEFdZc8wL+uAGiJSFVsS7Ak8enGm/UqVkMy2RZYCA3NLhJB7Mry7INEqpVRWhTW4qzEmQ0T6Yzvs5g18Z4zZJiLDgfXGmNn5qTfHZGiMOZ6/UJVS6nKFeXOJMWYuMDfbZ1e8XM8Y08qZOvUh8koplxPyN0SWO2kyVEq5nhTebrKraDJUSrmc7Q4UTYZKKeXxl6doMlRKuYWHdww1GSql3KFoxyp0hiZDpZTL6dlkpZSy0xMoLmYxhjPnM4o6jEz+Ja75RepyJ9a5/DlfVyW4y2dFHcJl4n9+vqhDyGTN98M+shB0N1kppXQ3WSml7LRnqJRS6HWGSimFAN7aM1RKKb3oWimlAEE8fEdZk6FSyi20Z6iUuuHZLq3x7GyoyVAp5XqiPUOllAL0djyllLIP7lrUUeROk6FSyi30bLJSSqHHDIvMkjU7eGfsTCxWwyOdm9D/8bYO8y+kZfDSe5PYvOsQwQF+fDGsN5XCy3Iw4RitHv+Q6pGhAETdXIUPBz4EwC+LYhg3cREClAsJZNzbj1MmqFShx75o5XbeGP0zFquVJ7o25eU+7Qu9DY0n/+6OqswHT7XE21uYuGAbn86IcZhfKbQ04wa0JSSwJCdOneeZTxYQf+x0gdtdsnoHb386E6vVyiP3NeGFJ9o5zL+QlsGAdyexZddBggP9+XK4bZu+6NCR47R6/ANefbIjzz3ahvMX0un2/GekpWeQkWGlU+sGDHrq3gLHmRNP7xl6+kAS+WKxWHnrk5+Z+PEzLJn4Or8uimX3P0ccykyds5rA0n6smPoW/36oFe9/+VvmvCoVyrLg+9dY8P1rmYkwI8PCO2NnMX3s8yz67/9Rp3oE389c7pLYB42cxvSx/Vg97S1mLIhh576EQm9H48kfLy9h1DOt6DHsV5o8P4nuLWtSq1IZhzLDn2zO1CU7aD7gf4z8aS1DejUtcLsWi5XBo6czefQzLJ38xhW36Sm/ryKodElWTnubfz/cihGf/+Ywf9i4X2jTpG7mdPFiPkz/rD+L/vt/LPzvayxds5OYrfsLHOuVXDxm6MyrqFyXyXDjjjiqVAihckQIxXx96Hr3rSz4a4tDmQXLt9Cjw20AdGrVgL9i/saYnAduM4AxhrPn0zDGcPrsecqFBBZ67DHb9lOtUghVKtpi79YuirnLNhd6OxpP/jSqUY59CSnEHT1JeoaVmcv/5t47qjmUqVWpDMs3HwJg+eZDdMw2Pz827IijSsVQKle4uE1HMX+54zY9f/lWetx7OwCdWzXgr5jdmdv0vOjNVAovS82q5TPLiwj+fsUBSM+wkJ5hcd2urAheTr6KisuToYhUEZGtWaYHishQEVkqImNFZKOIbBWR2+3z77J/tlFENohI6attMyEplfCw4Mzp8qFBJCSnOpQ5knypjI+PNwH+JTiRegaAAwnHuefJUXTvP441m/YC4Ovjzfuv9qBt749odP87/L3/CI90anLVy8OZ2CuUuxR7RLlgEpJSc/mGa2k8jsLLluJw8qVd3vjk04SX9Xcos+2fZDrfWR2AzndWJ8CvGMGlSxSo3SNJqUSEBV2KIyzost99JCmFiGzb9PHUM5w5e4HPJy3m1Sc7XFavxWKlbe+R3NL5TVreVouom6sUKM7ciJOvolLUPUM/Y0xDoB/wnf2zgcDz9s9bAOeyf0lEnhaR9SKy/lhycqEGFFY2kLU/v8P87wbxzgv303/4RE6dOU96hoWJv6zgj+8GEfPLMGpXj+A/kxYVatvq+vD293/RrF4Fln36CM1ursDh5NNYrNYii+fj7+bx74dbZfYCs/L29mLRf18jZtYwNm6PY+e+eJfEcPG5yZ7cMyzqEyhTAIwx0SISICJBwArgExGZDMw0xhzK/iVjzARgAkDDqEaX7duGhwaSkHgic/pIUgrh2XZpy4fYykSEBZGRYeHkmfMEB/ojIhQvZlsst9SqROWIsuw7mMjFPegqFUIAuK/1rYyfXPjJMDw0kMNHL8Uef/QE4aGFvzuu8eRPwrHTVAi5dNIsIqQUCcfOOJQ5cvwMvT6YC4B/CV/ua3oTJ8+kFajd8qGBxCemXIojMeWy310+NIj4bNt0mUB/NmyLY86STYz4fDYnT5/Dy76NP/lgy8zvBpb2o2lUDZas3kntahEFijUnnn36xD09w4xs7WTdX8ieyIwx5kPgKaAksEJEal9tgw1qR/LPoWQOxB8jLT2DXxdvoF3zeg5l2jWvx/Q/1gEwZ+kmmkXVQEQ4duI0Fovtr3hcfDL/HEomMqIs5UMD+Xv/UY6dsO0iLV+/ixqVy11taHmKqluZvQeSiDucTFp6BjMXxtKx5S2F3o7Gkz+xfx+lekQQkeUC8PXxoluLGsxbs8+hTJnSJTKPvb38YGMmL9pW4HYb1o7kn0NJWbbpWNpn26bbN6/H9LlrAfh96SaaN7Jt07988SJrZ7zD2hnv8NRDd/FCr3Y8+WBLjp04TeqpswCcu5BG9Lpd3FQ5rMCx5sjD95Pd0TM8CoSJSFngNNAZ+MM+72FgiYg0B1KNMakiUt0YswXYIiK3AbWBnVfToI+PN+++3J3HXv0Sq9XKw53uoFbVcEZ9M5cGtSNp37wePTs14cURk2jWcwRBAX58PrQXAKs37WX0t/Pw8fHCS7z4cGAPggNsx4Re7nsP3V/4DB9vbyqWL8OYwY8WxvK5LPaRrz1E9wHjsVgMj3VpQp3q4YXejsaTPxar4bWvljJjaFe8vbyYvGgbOw8e541H72DjnkTmrf2H5vUrMqRXU4wxrNwWz6Avlxa4XR8fb957uTuPvvIFFouVnp2bUKtaOCO/nkuD2pW4p0V9HunchAHvTqLpQ+8SZL9cLDdHj6Xy4ojJWK1WrFbDfW1upV2zerl+pyA8/XY8ye0MaqE1IjIAeBE4DOwD9gOtgI3AXYAv8KQxZq2IjANaA1ZgG9DHGHMhp7obRjUyi6PXuDT+q6FPx7v26NPxctem+R1siF1foExWp/6t5sdflzpV9vbqQTHGmMYFaS8/3PIv1xjzGeCwxYnIUmCSMealbGVfcEdMSik38+yOYZGfQFFK3QBshwM9OxsWWTI0xrQqqraVUm52DYxnWNTXGSqlbhCFeTJZRDqIyC4R2SMir19h/isisl1ENovIYhGpnFedmgyVUm4giDj3yrMmEW9gPNARqAs8IiJ1sxXbADQ2xtwC/AyMzKteTYZKKbcQce7lhNuBPcaYfcaYNGAq0DVrAWPMEmPMWfvkaqBiXpVqMlRKuZyzu8j2XBhy8XZb++vpbNVVAA5mmT5k/ywn/wLm5RWjnk1WSrmH8ydQkgvrOkMReRxojO165lxpMlRKuUUhXlpzGKiUZbqi/TPH9kTaAm8Cd+V248ZFupuslHKLQjxmuA6oISJVRaQY0BOY7diW3Ap8BXQxxiQ6U6n2DJVSrleI1xkaYzJEpD8wH/AGvjPGbBOR4cB6Y8xsYBRQCphuP0N9wBjTJbd6NRkqpdyiMO9AMcbMBeZm+2xIlvdtL/tSHjQZKqVcTvD8O1A0GSql3MLDc+G1nwy9RCjuq+eBVP4d+8XzBkoq2+K1og4h04Vdlw02nz8eng2v+WSolLo2ePrgrpoMlVJu4dmpUJOhUspdPDwbajJUSrmcDu6qlFJwTQzuqslQKeUWHp4LNRkqpdzBuYFbi5ImQ6WUW3h4LtRkqJRyvat5vklR0WSolHIPD8+GmgyVUm6hl9YopRR6zFAppUDAS5Oh+y1etZ03x8zEYrXyeJc7ebFXO4f5F9LSeX7YJDbtOkiZAH++HtGHyIiyHE89w5NvfMuGHQfo2ekOPhrYA4DTZ87T+dmxmd9PSEzhwQ6Nee/l7i6Jf9HK7bwx+mcsVitPdG3Ky33au6QdjcfR4lXbeeOTGVjt281LvR3rvZCWTr9hE9m08yDBgf58O6IvkRFlARjzwwIm/7YKLy8vPnz1Qdo0qQPAF1P+ZOKvqxAR6lYPZ9zbj1OiuC+dnh7D6bO2x3IknThFVN3KTBqV/SFwObv7jlp88FIXvL28mPjbWj6dtMRhfqVyQYwb/BAhQaU4cfIszwyfQnxSKs2jqvP+gEsDPteIDOVf70xm7vJt+VpmV8ezs6FHjH0lIvtFJEREqojI1oLUZbFYef3j6Uwd8ywrpgxm1oIYdv2T4FBm8uzVBAX4se7nITz7SCuGj7c9PqF4MR9ef7oTw16436F8Kf8SLJ34f5mviuWD6dSqQUHCzDX+QSOnMX1sP1ZPe4sZC2LYuS8h7y+6yI0Sj8Vi5bVR05n26XOsnPomM69Q76TZqwgq7cf6Ge/wXM/WDBv/KwA79yUwa2EMK6YMZvrY5xg0choWi5X4xBQm/LSMxT8MYsWUwVishpkLYwCYM+Fllk16nWWTXue2elXp3Nr57cnLSxj16gP0ePVbmjz2Md3bNqRWlTCHMsP7d2bqHzE07/0JI79fyJBnOwLwV+xeWvYZQ8s+Y+jywpecu5DOkrW7C7LonHJxcNdCegaKS3hEMixMsdvjqFIxlCoVQijm68P97aKYF73Focy85Vt4+N7bAbivdUOWr9+NMQb/ksVp0rA6xYv55lj/3gOJJJ84zZ0Nq7sk/pht+6lWKYQqFW3xd2sXxdxlm13SlsZzSez2OKpWDMncbh5o1+jy7SZ6Cz073QFAlzYNiV5n227mRW/hgXaNKF7Ml8oRIVStGELs9jgAMixWzl9IJyPDwrnzaYSHBDrUefL0OZbH7Obelrc4HWujOpHsO5RMXPxx0jMszFy8kXtb3OxQplbVciyP2QPA8ti9dMw2H6Br61tYtHon5y6kO912QVzFc5OLhFuSoYgMEpEB9vdjRORP+/s2IjI5h+9UE5ENInLb1bSVkJRChbCgzOmIsCASklIdyhxJSqVCOVsZHx9vAkqV4HjqGafqn7UwhvvbRrnsavqEpFQqlAvOnI4oF3xZ/O50o8STkJjiWG9YEAlJKZe1HRGWdbspyfHUM7ZtLvt3E1OICAui/2N306DrEOp2eouAUiVpbd99vmhu9GZaNq5FQKmSTscaHhrA4cRLscUnphIe6phkt/2dQOe76gPQ+a56BPiXIDjAz6FMt7YNmbFwo9PtFpT2DG2WAy3s7xsDpUTE1/5ZdPbCIlILmAH0Mcasu8L8p0VkvYisT05KcmHYl5u1MJZu7aPc2qa6NqWcPMvc6M3EzhrKtjkjOHPuAtPmOW7OMxfE0K19o0Jv++3xv9Ps1mos+/4lmjWsxuHEFCxWa+b8cmVLU7daeRav2VXobedERJx6FRV3JcMYoJGIBAAXgFXYkmILbIkyq1DgV+AxY8ymK1VmjJlgjGlsjGkcEhrqMC88NCjbX82Uy/5qlg8N5PBRW5mMDAsnT5+nTKB/nj9i69+HybBYaVA7Ms+y+RUeGsjhoycyp+OPnrgsfne6UeIJDwtyrDcxhfDQIMcyoYHEJ2bdbs5RJtDfts1l/25YEMvW7aJyRFlCgkvj6+NN59YNWLtlX2a5Yymnid0WR/tml+/C5iYh6WS2vZ/Ay/d+kk/Sa/CP3NX3U0ZM+AOAk6fPZ86/v00Dfo/eSobFirvobjJgjEkH/gH6ACuxJcDWwE3AjmzFU4EDQPP8tHVrnUj+OZhEXPwx0tIz+GVhLB1a1Hco06FFPX6auxaA35ZspHnjGk79RXLVX/GsoupWZu+BJOIOJ5OWnsHMhbF0vIrjSRpP/txaJ5J9B5OIi7fVO2thDB1bZt9u6jN1zhoAZv+5kRaNayIidGxZn1kLY7iQlk5cfDL7DiYRVbcyFcoFs37rfs6eT8MYQ/S63dSsUj6zvtl/bqR983qUKJ7zMeorid15kOoVQ4gMD8bXx5tudzdk3l/bHcqUCfTL3KZffqINk+c49ki7t2vIjEWet4tclLvJ7ry0ZjkwEHgS2AJ8AsQYY0y2RJQGPADMF5HTxpj/XU0jPj7efDDwQR568XOsViuPdG5C7WrhfDhhDg1rR9KhZX0eu+9O+g2byG0PDic4wI8J7/bJ/H7U/UM5dfY8aekZzFu2memf9aNW1XAAZi/ewJRPni3AInAu/pGvPUT3AeOxWAyPdWlCnerhLm1T47HV+9HAHvQY8DkWq+HR+2zbzQdfzaFhnUg6tqzP413u5LmhP9K4+zCCAvz4ZkRfAGpXC6dr2yia9nwfb28vRg7qgbe3F43rVaFLm4a07vURPt7e1K9Zkd73N81sc9bCmMsu+3KGxWLltTG/MOOTf+Pt7cXk39ey85+jvPFUezbuPMS8v7bT/NbqDHm2I8bAyk37GDR6Vub3K5UPpkJYECs27MullcLn6XegiDHGPQ2J3A38AQQZY86IyG7gS2PMJyKyH/uxROB3Y0w9EQkCFgLvGmNm51RvVKPGJnrlWjf8Auf4eF93J+ive1are/4NXA2Pejre1olYzxwpUCZrGNXILIxe41TZsNK+McaYxgVpLz/c1jM0xiwGfLNM18zyvor9bTJQz/5ZCnBVZ5KVUp7Ls/uF1+kdKEopTyP6qFCllLp4B4on0wNcSimF9gyVUm7i6T1DTYZKKbfw9EtrNBkqpVxPn5uslFLXxgkUTYZKKbfQ3WSllMLze4Z6aY1Syi0Kc9QaEekgIrtEZI+IvH6F+cVF5Cf7/DUiUiWvOjUZKqXco5CyoYh4A+OBjkBd4BERqZut2L+AE8aYm4AxwEd51avJUCnlcgJ4iTj1csLtwB5jzD5jTBowFeiarUxX4L/29z8Dd0se4/Rd88cMN8TGJJcu4R1XCFWFYBsowlN4WjzgeTFpPLkrrHgqF7SC2NiY+SV9JcTJ4iVEZH2W6QnGmAlZpisAB7NMHwLuyFZHZhljTIaIpAJlyWV5XPPJ0BgTmnepvInI+qIYNignnhYPeF5MGk/uPCkeY0yHoo4hL7qbrJS61hwGKmWZrmj/7IplRMQHCASO5VapJkOl1LVmHVBDRKqKSDGgJ5B9AOjZQG/7+weBP00eI1lf87vJhWhC3kXcytPiAc+LSePJnafFUyjsxwD7A/MBb+A7Y8w2ERkOrLePjP8tMFFE9gDHsSXMXLlt2H+llPJkupuslFJoMlRKKeAGTYYiMlREBhZ1HBeJyMqijiEvIvLNxav8RWRwEcfykoj45TCvj4j8x90xeSIR2S8iISJSRUS2FnU8nu6GTIaexhjTNO9SRcsY85Qx5uKTyos0GQIvAVdMhkrl1w2TDEXkTRHZLSJ/AbUALxGJsc9rICJGRCLt03tFxE9EeojIVhHZJCLRLozttNiMsre3RUQets/zEpHPRWSniCwUkbki8qCrYrG36S8ic+y/e6uIPCwiS0WksYh8CJQUkY0iMtmVceQQyztABLBERJbYy/S1r9u1QDMXxuLQwxKRgfa9jKUiMta+TLaKyO32+XfZP9soIhtEpHQhxzNIRAbY348RkT/t79vktG5EpJo9Fn0MbzY3xKU1ItII26n1hth+cywQg+22nwCgBbAeaGFPlonGmLMiMgS4xxhz2P5Qe1fqZo+vAbbbqNbZE3AzoAq2G9LDgB3Ady6OpQMQb4zpBCAigcBzAMaY10WkvzGmoYtjyC2WvkBrY0yyiIQDw4BGQCqwBNjgptiy8jPGNBSRltjWTz1gIPC8MWaFiJQCzhdym8uBV4HPgMZAcRHxxbY9X9x2MolILWz38fYxxmwq5FiueTdKz7AFMMsYc9YYc5JLF2iuxLbBtATet/+/BbaNDGAF8IOI/Bvb9Uyu1ByYYoyxGGOOAsuA2+yfTzfGWI0xR7D9Y3e1LUA7EflIRFoYY1Ld0GZ+Y7kDWGqMSbLftP+T+0MEYAqAMSYaCLD/8VwBfGLvvQUZYzIKuc0YoJH9D/oFYBW2pJh1G74oFPgVeEwT4ZXdKMkwJ9HYNpzK2DaUBtiSz3L4//bOL8SqKgrjvw/JlMY0JA0rsEbFyP6gGJlRWjKUWRghYYmIpo1BJJERZFEkZNhTD1HWg4QQFSaEYSCCOsmo2dhYVipl+dC/eaghR6VJvx72HjlznaujzW1s7vrBMGfOstLkRwAABMhJREFU2XevdS7nrrv2mrO+A7brgWWktp7PJQ3tJT//U2zvB8aTAtHynCFXvS/A33T+zAwobJfesGvbK4BHgIHANklje9IZ2+3AQWAe6Yu9AZgKjCKtIIq0AodI13fQBdUSDLcCMyUNzHWbe/P+BmAOcMD2CdKd6tOBTwEk1dreYft5oIXO/ZA9TQPwoKR+ki4lZak7SdnFA7l2OByYUkEfAJA0Ajhiew2wkhSMirTn5VjFKePLn0BH/W0HcLukodmnWRV051dgWLZ1ITCjcKyjxnsr0Gq7NV8/X9p+hdRC1qPBMNNAWo5vzdv1wO4uWs/+Au4H5kp6qAJ+/O+pipqh7SZJ7wHNwG+kCxPbP0gS6UKCFASvsP17/nulpNEkObZN+fUVcRFYB0zKNgw8bfsXSWuBO4GvSZJETaRv+UpyHencTwDtpHrhq4Xjq4A9kppsP9wLvkwCPpH0k+2pkl4gLRH/AL6olCO225VavnaShAC+LRw+Jmk3cAEwP+9bImkqcALYC2yogFsNwLNAo+02Scc4dYnc4X+bpBnARkmHc9takIl2vF4mL72bbJfVjJNUY/twHrsTmJzrh8F5gKTNwFO2d51pbHD+UhWZ4flKXgJupnPW1RXrc0G+P/BSBMIg6HkiMwyCIKB6/oESBEFwWiIYBkEQEMEwCIIAiGDY55F0vNAz+4HKqL10c64pktbn7fvUxcO7C2OHSHrsHGx0qShUbn/JmNVn07dd2mscVDcRDPs+R23faHsc6cbb+uJBJc76OrD9Ue6wKMcQ4KyDYRD0FhEMq4sGYFTOiPZJegf4CrhSUp2kRklNOYOsAZB0l5JiThNJTIK8/6RuoKThktYpKcs0S7oFWAHU5qx0ZR63VNJnkvZIerEwV6mi0GmRtDDP0yxpbUm2O03SrjzfjDy+n5IiUIftR//tGxn0PSIYVglKj0u8m9TjCzAaeN32tUAbqQd7mu3xJAWfJyUNAN4itS9OAC4rM/1rwBbbN5Da5fYCzwDf5ax0qaS6bPMmkjrPBEm3qbOi0HSSOMWZ+ND2xGzvG2BB4djIbOMe4I18DgtILXIT8/wLJV3VDTtBFRE3Xfd9BkrqaFFrID01bATwo+3tef/NJImwbak7kf6k9raxwEHbBwAkrQEWdWHjDmAugO3jQKukS0rG1OWfDnmtGlJwHERWFMo2utMiNk7SctJSvIb0lLQO3s995gckfZ/PoQ64vlBPHJxt7++GraBKiGDY9zlaqj2YA15bcRew0fbsknE9qVko4GXbb5bYWHIOc60GZtpuljSPzuIVp6jHZNuP2y4GTSSNPAfbQR8llskBwHZgsqRRcFJdegxJiGCkpNo8bnaZ128ii7/m+txgOivLQMre5hdqkZdLGkZ5RaHTMQj4OavUlApFzFJS+KkFrgb2ZduL83gkjZF0UTfsBFVEZIYBtltyhvWukjQVwDLb+yUtAj6WdIS0zO5Kuv4JYJWkBcBxYLHtRknb8q0rG3Ld8BqgMWemh4E55RSFzsBzJOmulvy76NMhkpjFxUC97WOS3ibVEpuUjLcAM7v37gTVQvQmB0EQEMvkIAgCIIJhEAQBEMEwCIIAiGAYBEEARDAMgiAAIhgGQRAAEQyDIAgA+Ae+qrevOGvg6wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(grid_search, X_test, y_test,\n",
    "                      display_labels=label_encoder.classes_,\n",
    "                      cmap=plt.cm.Blues,\n",
    "                      normalize=\"true\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Convolutional Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Activation, Flatten, SpatialDropout1D, Conv1D\n",
    "from keras.wrappers.scikit_learn import KerasClassifier\n",
    "from keras.utils import to_categorical"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_model(optimizer=\"adam\", dropout=0.1, nbr_features=2500, dense_nparams=256):\n",
    "    model = Sequential()\n",
    "    model.add(Dense(dense_nparams, activation='relu', input_shape=(nbr_features,))) \n",
    "    model.add(Dropout(dropout), )\n",
    "    model.add(Dense(6, activation='sigmoid'))\n",
    "    model.compile(loss='categorical_crossentropy', optimizer=optimizer, metrics=[\"accuracy\"])\n",
    "    return model"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  },
  "pycharm": {
   "stem_cell": {
    "cell_type": "raw",
    "metadata": {
     "collapsed": false
    },
    "source": []
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}