irc_model_linreg.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Linear Regression Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/Cellar/python/3.7.4/Frameworks/Python.framework/Versions/3.7/lib/python3.7/importlib/_bootstrap.py:219: RuntimeWarning: numpy.ufunc size changed, may indicate binary incompatibility. Expected 216, got 192\n",
      "  return f(*args, **kwds)\n",
      "/usr/local/Cellar/python/3.7.4/Frameworks/Python.framework/Versions/3.7/lib/python3.7/importlib/_bootstrap.py:219: RuntimeWarning: numpy.ufunc size changed, may indicate binary incompatibility. Expected 192 from C header, got 216 from PyObject\n",
      "  return f(*args, **kwds)\n"
     ]
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "from irc_utils import compute_score, exponential_moving_average\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV\n",
    "from sklearn.metrics import classification_report, confusion_matrix, accuracy_score, f1_score\n",
    "from sklearn.preprocessing import StandardScaler"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " #### Data processing\n",
    "- Each row in csv file represent one session\n",
    "- Each session is represented by\n",
    "    - periodicity, duration, total packet size, message count, number of source ports, destination port"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>periodicity</th>\n",
       "      <th>duration</th>\n",
       "      <th>pkt_size</th>\n",
       "      <th>msg_count</th>\n",
       "      <th>src_ports_count</th>\n",
       "      <th>dst_port</th>\n",
       "      <th>src_spec_chars</th>\n",
       "      <th>msg_spec_chars</th>\n",
       "      <th>msg_word_entropy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>622.000000</td>\n",
       "      <td>622.000000</td>\n",
       "      <td>622.000000</td>\n",
       "      <td>622.000000</td>\n",
       "      <td>622.0</td>\n",
       "      <td>622.000000</td>\n",
       "      <td>622.000000</td>\n",
       "      <td>622.000000</td>\n",
       "      <td>622.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.284318</td>\n",
       "      <td>155264.967846</td>\n",
       "      <td>26300.752412</td>\n",
       "      <td>89.705788</td>\n",
       "      <td>1.0</td>\n",
       "      <td>45767.950161</td>\n",
       "      <td>0.245198</td>\n",
       "      <td>0.253180</td>\n",
       "      <td>6.280716</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.338709</td>\n",
       "      <td>130326.750331</td>\n",
       "      <td>60053.403822</td>\n",
       "      <td>204.684782</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10969.030234</td>\n",
       "      <td>0.126221</td>\n",
       "      <td>0.140154</td>\n",
       "      <td>2.345869</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.002800</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>360.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2994.000000</td>\n",
       "      <td>0.095200</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.043625</td>\n",
       "      <td>6327.500000</td>\n",
       "      <td>1267.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>38060.000000</td>\n",
       "      <td>0.157900</td>\n",
       "      <td>0.191100</td>\n",
       "      <td>4.752650</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.123400</td>\n",
       "      <td>169951.000000</td>\n",
       "      <td>3795.000000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>45959.000000</td>\n",
       "      <td>0.214300</td>\n",
       "      <td>0.228150</td>\n",
       "      <td>6.407250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.322525</td>\n",
       "      <td>288203.000000</td>\n",
       "      <td>20638.000000</td>\n",
       "      <td>93.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>52790.000000</td>\n",
       "      <td>0.272700</td>\n",
       "      <td>0.277800</td>\n",
       "      <td>8.011800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>619200.000000</td>\n",
       "      <td>713848.000000</td>\n",
       "      <td>4099.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>65124.000000</td>\n",
       "      <td>0.833300</td>\n",
       "      <td>0.909100</td>\n",
       "      <td>10.707400</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       periodicity       duration       pkt_size    msg_count  \\\n",
       "count   622.000000     622.000000     622.000000   622.000000   \n",
       "mean      0.284318  155264.967846   26300.752412    89.705788   \n",
       "std       0.338709  130326.750331   60053.403822   204.684782   \n",
       "min       0.002800       1.000000     360.000000     3.000000   \n",
       "25%       0.043625    6327.500000    1267.000000     8.000000   \n",
       "50%       0.123400  169951.000000    3795.000000    23.000000   \n",
       "75%       0.322525  288203.000000   20638.000000    93.000000   \n",
       "max       1.000000  619200.000000  713848.000000  4099.000000   \n",
       "\n",
       "       src_ports_count      dst_port  src_spec_chars  msg_spec_chars  \\\n",
       "count            622.0    622.000000      622.000000      622.000000   \n",
       "mean               1.0  45767.950161        0.245198        0.253180   \n",
       "std                0.0  10969.030234        0.126221        0.140154   \n",
       "min                1.0   2994.000000        0.095200        0.000000   \n",
       "25%                1.0  38060.000000        0.157900        0.191100   \n",
       "50%                1.0  45959.000000        0.214300        0.228150   \n",
       "75%                1.0  52790.000000        0.272700        0.277800   \n",
       "max                1.0  65124.000000        0.833300        0.909100   \n",
       "\n",
       "       msg_word_entropy  \n",
       "count        622.000000  \n",
       "mean           6.280716  \n",
       "std            2.345869  \n",
       "min           -0.000000  \n",
       "25%            4.752650  \n",
       "50%            6.407250  \n",
       "75%            8.011800  \n",
       "max           10.707400  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Select model version\n",
    "# model_version = '1' \n",
    "model_version = '2'\n",
    "\n",
    "plt.rcParams[\"figure.figsize\"] = (10,7)\n",
    "\n",
    "# Import data\n",
    "data = pd.read_csv('input/trn_data_v{}.csv'.format(model_version), delimiter=';')\n",
    "\n",
    "# Prepare data for training\n",
    "# Divide the data into attributes and labels\n",
    "X = data.iloc[:, 1:-1]\n",
    "y = data.iloc[:, -1].values\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.4, random_state=0)\n",
    "\n",
    "# showing examples of data\n",
    "X_train.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1037, 9)"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sc = StandardScaler()\n",
    "X_train = sc.fit_transform(X_train)\n",
    "X_test = sc.transform(X_test)\n",
    "X = sc.transform(X)\n",
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SGDClassifier(alpha=0.0001, average=False, class_weight=None,\n",
       "              early_stopping=False, epsilon=0.1, eta0=0.0, fit_intercept=True,\n",
       "              l1_ratio=0.15, learning_rate='optimal', loss='hinge',\n",
       "              max_iter=1000, n_iter_no_change=5, n_jobs=None, penalty='l2',\n",
       "              power_t=0.5, random_state=None, shuffle=True, tol=0.001,\n",
       "              validation_fraction=0.1, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import SGDClassifier\n",
    "# C, kernel, gamma = clf.best_params_['C'], clf.best_params_['kernel'], clf.best_params_['gamma']\n",
    "linreg = SGDClassifier()\n",
    "\n",
    "linreg.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix:\n",
      "[[39  0]\n",
      " [ 0  3]]\n",
      "\n",
      "Sensitivity(=Recall) TPR = TP / (TP + FN):\t\t1.0\n",
      "Specificity SPC = TN / (FP + TN):\t\t\t1.0\n",
      "Precision PPV = TP / (TP + FP):\t\t\t\t1.0\n",
      "Negative Predictive Value NPV = TN / (TN + FN):\t\t1.0\n",
      "False Positive Rate FPR = FP / (FP + TN)):\t\t0.0\n",
      "False Discovery rate FDR = FP / (FP + TP):\t\t0.0\n",
      "False Negative rate FNR = FN / (FN + TP):\t\t0.0\n",
      "Accuraccy ACC = (TP + TN) / (P + N):\t\t\t1.0\n",
      "F1-score F1 = 2TP / (2TP + FP + FN):\t\t\t1.0\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[1.0, 1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 1.0, 1.0]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from irc_utils import compute_score\n",
    "\n",
    "y_pred = linreg.predict(X_test)\n",
    "\n",
    "compute_score(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Improvement\n",
    "\n",
    "#### Higher number of samples?\n",
    "- could help in case we don't have overfitted model\n",
    "- To show the accuracy plot, Random Forrest classifier with tuned hyperparameters will be trained multiple times with different training dataset size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:8: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  TPR = np.nan_to_num(round(TP / (TP + FN), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:10: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  PPV = np.nan_to_num(round(TP / (TP + FP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:13: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FDR = np.nan_to_num(round(FP / (FP + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:14: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FNR = np.nan_to_num(round(FN / (FN + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:16: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  F1 = np.nan_to_num(round(2*TP / (2*TP + FP + FN), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:8: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  TPR = np.nan_to_num(round(TP / (TP + FN), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:10: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  PPV = np.nan_to_num(round(TP / (TP + FP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:13: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FDR = np.nan_to_num(round(FP / (FP + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:14: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FNR = np.nan_to_num(round(FN / (FN + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:16: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  F1 = np.nan_to_num(round(2*TP / (2*TP + FP + FN), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:8: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  TPR = np.nan_to_num(round(TP / (TP + FN), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:10: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  PPV = np.nan_to_num(round(TP / (TP + FP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:13: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FDR = np.nan_to_num(round(FP / (FP + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:14: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FNR = np.nan_to_num(round(FN / (FN + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:16: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  F1 = np.nan_to_num(round(2*TP / (2*TP + FP + FN), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:10: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  PPV = np.nan_to_num(round(TP / (TP + FP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:13: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FDR = np.nan_to_num(round(FP / (FP + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:8: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  TPR = np.nan_to_num(round(TP / (TP + FN), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:10: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  PPV = np.nan_to_num(round(TP / (TP + FP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:13: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FDR = np.nan_to_num(round(FP / (FP + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:14: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FNR = np.nan_to_num(round(FN / (FN + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:16: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  F1 = np.nan_to_num(round(2*TP / (2*TP + FP + FN), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:8: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  TPR = np.nan_to_num(round(TP / (TP + FN), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:10: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  PPV = np.nan_to_num(round(TP / (TP + FP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:13: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FDR = np.nan_to_num(round(FP / (FP + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:14: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FNR = np.nan_to_num(round(FN / (FN + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:16: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  F1 = np.nan_to_num(round(2*TP / (2*TP + FP + FN), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:8: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  TPR = np.nan_to_num(round(TP / (TP + FN), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:10: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  PPV = np.nan_to_num(round(TP / (TP + FP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:13: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FDR = np.nan_to_num(round(FP / (FP + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:14: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FNR = np.nan_to_num(round(FN / (FN + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:16: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  F1 = np.nan_to_num(round(2*TP / (2*TP + FP + FN), 4))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "shape: 80,9\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:8: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  TPR = np.nan_to_num(round(TP / (TP + FN), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:10: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  PPV = np.nan_to_num(round(TP / (TP + FP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:13: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FDR = np.nan_to_num(round(FP / (FP + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:14: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  FNR = np.nan_to_num(round(FN / (FN + TP), 4))\n",
      "/Users/preneond/Documents/Work/Stratosphere/Masters-thesis/IRC-Malware-Detection/irc_utils.py:16: RuntimeWarning: invalid value encountered in long_scalars\n",
      "  F1 = np.nan_to_num(round(2*TP / (2*TP + FP + FN), 4))\n"
     ]
    }
   ],
   "source": [
    "train_ratio = np.arange(0.2,1,0.01)\n",
    "n_samples = len(X)\n",
    "train_size = list(map(lambda x: round(x*n_samples), train_ratio))\n",
    "\n",
    "score_trn_list = []\n",
    "score_tst_list = []\n",
    "for trn_ratio in train_ratio:\n",
    "    X_tmp, _, y_tmp, _ = train_test_split(X, y, train_size=trn_ratio, random_state=0)\n",
    "    X_train, X_test, y_train, y_test = train_test_split(X_tmp, y_tmp, train_size=0.8, random_state=0)\n",
    "    logreg = SGDClassifier()\n",
    "    logreg.fit(X_train, y_train)\n",
    "    \n",
    "    y_pred_trn = logreg.predict(X_train)\n",
    "    y_pred_tst = logreg.predict(X_test)\n",
    "    \n",
    "    score_trn = compute_score(y_train, y_pred_trn, verbose=False)\n",
    "    score_tst = compute_score(y_test, y_pred_tst, verbose=False)\n",
    "    score_trn_list.append(score_trn)\n",
    "    score_tst_list.append(score_tst)\n",
    "\n",
    "print('shape: {},{}'.format(len(score_trn_list), len(score_trn_list[0])))\n",
    "score_trn_list = np.stack(score_trn_list, axis=0)\n",
    "score_tst_list = np.stack(score_tst_list, axis=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Sensitivity(=Recall) TPR = TP / (TP + FN):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmEAAAG5CAYAAADGcOOUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdd3zV1f3H8dfJIoQkBJKQAAHC3hAlIiAqggwBceLCWlepba3Vavuz1qpV21pb67YWFVdV1LpRwYETZcvesncSIIwkZJ3fHycJCSQh4958b5L38/G4jzu+3/u9n1zIzeee8/l+jrHWIiIiIiJ1K8jrAEREREQaIyVhIiIiIh5QEiYiIiLiASVhIiIiIh5QEiYiIiLiASVhIiIiIh5QEiYiEkCMMcnGGGuMCfE6FhHxLyVhIlIrxphDpS6FxpjsUvcnGWPuMcbkFd3fb4z5zhgzuOi5VxtjCoq2HTDGLDHGjPf6ZxIRqQtKwkSkVqy1kcUXYAtwbqnHXina7fWi7fHAt8DbxhhTtO37om0xwFPANGNMjL/j1kiTiHhNSZiI1BlrbR7wIpAIxB6zrRB4GWgGdC3v+caYYcaYbcaYO4wx6caYTcaYSaW2NzHG/NMYs8UYs9sY87Qxpukxz/0/Y8wu4Plyjt/FGPOVMSaz6Pivl9r2qDFma9GI3UJjzOmltt1jjHnTGPNfY8xBY8wyY0w3Y8wfjDF7ip43qtT+Xxpj/maMmVd0vPeMMS0r+JmbG2OeM8bsNMZsN8bcb4wJPlG8IhL4lISJSJ0xxjQBrga2WmvTj9kWDFwD5AGbKzlMIhAHtAV+CkwxxnQv2vYA0A1IAboU7XPXMc9tCXQAJpdz7PuAT4AWQBLweKlt84uO2xJ4FXjTGBNeavu5uCSyBfADMBP3GdsWuBf4zzGvdRVwLdAayAceq+DnfaFoexfgJGAUcH0V4hWRAKckTETqwiXGmP3AVmAAcEGpbYOKtuUA/wSutNbuOcHx/mStPWKt/Qr4sOj4BpdY3WKt3WutPQj8Fbis1PMKgbuLnptdznHzcAlaG2ttjrX22+IN1tr/WmszrLX51tqHgCZA91LP/cZaO9Namw+8iZt6faBo9G8akHzMNOvL1trl1trDwJ+Kfobg0sEYYxKAscDN1trDRe/Lw6V+pgrjFZHApyRMROrCG9baGGttK2vtcGvtwlLb5lhrY3CjOe8Dp5d/iBL7ihKXYpuBNrikJwJYWHQCwH5gRtHjxdKstTmVHPv3gAHmGWNWGGOuLd5gjLnNGLOqaOpvP9AcNyJXbHep29lAurW2oNR9gMhS+2w95mcIPeZ44BKsUGBnqZ/pP0CrE8UrIoFPhakiEhCstYeMMb8ANhhjplprf6hg1xbGmGalErH2wHIgHZfs9LbWbq/oZU4Qwy7gZwDGmKHAZ8aYr3FThr8HRgArrLWFxph9uASoptqVut0eN6qVfszjW4EjQFzRCFuV4rXWrq9FXCJSRzQSJiIBw1q7F3iWsnVc5fmzMSasqDh+PPBmUWH/M8DDxphWAMaYtsaY0VV9fWPMRGNMUtHdfbikrRCIwtVlpQEhxpi7gOhq/GjludIY08sYE4GrGftfqZEzAKy1O3E1Xw8ZY6KNMUHGmM7GmDNPEK+I1ANKwkQk0DwCjDXG9Ktg+y5cwrEDeAW4wVq7umjb/wHrgTnGmAPAZ5St2zqRU4C5xphDuKnR31hrN+CK7GcAa3FThzmUnU6siZdxRfe7gHDgpgr2uwoIA1bifu7/4UbmKotXROoBY22lo/MiIgHDGDMM+K+1NulE+wYyY8yXuJ/jWa9jERHvaCRMRERExANKwkREREQ8oOlIEREREQ9oJExERETEA/WuT1hcXJxNTk72OgwRERGRE1q4cGG6tTa+vG31LglLTk5mwYIFXochIiIickLGmArXwtV0pIiIiIgHlISJiIiIeEBJmIiIiIgH6l1NmIiIiPhOXl4e27ZtIycnx+tQ6rXw8HCSkpIIDQ2t8nOUhImIiDRi27ZtIyoqiuTkZIwxXodTL1lrycjIYNu2bXTs2LHKz9N0pIiISCOWk5NDbGysErBaMMYQGxtb7dFEJWEiIiKNnBKw2qvJe6gkTERERMQDSsJERETEExkZGaSkpJCSkkJiYiJt27YtuZ+bm1ulY1xzzTWsWbOmSvs+++yzJccPCwujb9++pKSk8Mc//pFnn32W+Ph4UlJS6NmzJ1OnTi15TvHjPXr04LHHHqvxz3usereAd2pqqlXHfBEREd9YtWoVPXv29DoM7rnnHiIjI7ntttvKPG6txVpLUJBvx42SkpJYvnw5MTExgEu2li9fziOPPMKuXbvo06cPq1ev5t133y15PC0tje7du7NixQpat2593DHLey+NMQuttanlxaCRMBEREQko69evp1evXkyaNInevXuzc+dOJk+eTGpqKr179+bee+8t2Xfo0KEsXryY/Px8YmJiuP322+nfvz+DBw9mz549NXr9xMREkpOT2bJlS5nH4+Pj6dSpEzt37qzVz1dMLSpEREQEgJtvnsHixbt8esyUlEQeeWRMtZ+3evVqXnrpJVJT3SDSAw88QMuWLcnPz+ess87i4osvplevXmWek5mZyZlnnskDDzzAb3/7W6ZOncrtt99e7ddev349mzdvplOnTixatKjk8U2bNlFQUECfPn2qfczyaCRMREREAk7nzp1LEjCA1157jZNPPpmTTz6ZVatWsXLlyuOe07RpU8455xwABgwYwKZNm6r1mq+88gopKSlMmjSJZ599tmSq8pVXXqFv37507dqVX//614SFhdX8ByvFbyNhxpipwHhgj7X2uJTRuHM5HwXGAlnA1dbaRcfuJyIiInWjJiNW/tKsWbOS2+vWrePRRx9l3rx5xMTEcOWVV5bbk6t0chQcHEx+fn61XnPSpEk88sgjFT4+d+5czjnnHMaPH0+rVq2qdezy+HMk7AWgsn/Nc4CuRZfJwL/9GIuIiIjUUwcOHCAqKoro6Gh27tzJzJkzPYnj1FNP5fLLL+fxxx/3yfH8NhJmrf3aGJNcyS7nAS9Zd3rmHGNMjDGmtbXWN9VuNbRjzQo2fvtBnbxWrolmZ5Oh0ACa5AUFGUaP7kyLFk29DkVERBqYk08+mV69etGjRw86dOjAaaed5lkst99+OwMHDuT2228vM1pXE35tUVGUhE2vYDpyOvCAtfbbovufA/9nrT2u/4QxZjJutIz27dsP2Lx5s99invvKE5y669d+O/6x+j90A0t3JtbZ6/lTjx5xfP75VbRpE+V1KI1WQUEheXmF5OUVHHedm3v8Y/n5bv/8/KOX4sereikosOTnF9KnTyvGj+9GTEy412+Dz5V+X0u/j8feLn4/S7/H4eEhDBqURLNmvqkhEfG1QGlR0RBUt0VFvTg70lo7BZgCrk+YP1+r99gr2LjxFH++BACh2dtImnMxM17uy8F2l/r99fxt3boMLrvsLc488wVmzbqKdu2aex1SnbLWkptbQHZ2PllZeSWX7Ozi63xycvLJzna3s7PzyMnJ58iRgqJrd7v4unjbkSP55OYWlNlW/Fjx4+46n7y8QgoL677vX3CwwRhDfn4hoaFBjBjRiQsv7MF55/WgVavafUvMzy8s854Vv7/HPlb8flZ0Ofo+lX1Pi9+70reLt5VOsmr7voaFBXPGGR0YM6YzY8Z0oVev+BovE1P6/1rp/1/h4SF07txCy8+I1CNeJmHbgXal7icVPeapyBYtiWxxqv9fqHAAzA+jdcgmWneL9f/r+Vm3brF8+ulPGD36v0WJ2E9JTo7xOqzjWGvJzs4nMzOHgwdzOXjwCIcO5Za5HDx49Pbhw7kcOpRXZnvp5Kr4D2BWVl6N/1AbA+HhITRpEkKTJsEl18c+FhUVVnI/LMxdim+HhhZfB5XcDw0NKnNdvP3YbSEh7hIaGlxy290PKnM/JCSI4GD3eHBw8X2XgBUWWubO3cbbb6/i7bdXM3nydG644UNOP709EyZ0JyoqrMz76t73vJL3+PDhvKL32t0ufo/z8gpr/G8dFlb8HgaXed9Kv6cREaG0aBFe8njxe1r8Ph773lb0PoeFBZe8h8e+x/v2ZfPppxuYMWM9t932Kbfd9ilJSdGMGdOZ/v0TycrKO+H7Uvo6OzuPiiYweveOZ9KkvlxxRV86dAi83z8RKcvL6chxwI24syNPBR6z1g480TEbVMf8l1Igsg1c+JHXkfjMggU7GDnyZaKjmzBr1lV07tzSp8cvTqL27ctm374c9u/PYd++7KJrd//Y2/v355CZmcOBA0fIzDxCfn7V/rCHhQUTGRlW5hIREUqzZqFERLhL06YhpW4f/3jTpu72sdfh4e52kybuj3dDGr2w1rJ06e6ShGz58rLNEps2DTnufW3WLIxmzUJLrqvyHlb0njZt6pLXoKDAe0+3bs1k5swfmTFjPZ99toHMzCOAS8SL34uoqCZl3ouj1+528Xtz7M+/e/dhpk1bzuzZWwE4/fT2TJrUl4kTe9OyZcOu1czNLSAoyBASoq5LNaHpSN+p7nSk35IwY8xrwDAgDtgN3A2EAlhrny5qUfEE7gzKLOCa8urBjtWgkrCProRt38Bk/9W4eeGHH3YycuTLhIeHMGvWT+lWyUiftZZ9+3LYvfsQe/YcZvfuw+zZc5i0tMOkpWWRnu4uxbf37s0mN7eg0tePjAwjJiacFi3CiYlxl+bNw4mODqN583CaN29C8+bhREW5P3hH//iVTQrCwoJ9/dY0Sjt2HMRaW/LeBgfrDyW4qdaMjKyS5N5XifjGjft49dVlvPLKMlatSicsLJhHHx3DDTeU+zcg4FhrOXw4jz17Dpdc0tKO3k5Pzy75XEhPzyIjI4uDB3OJigpj1KjOjBvXlXPO6UpiYmS1Xrew0LJ3b3bJZ0/paxeD+xxKSGjGJZf0ZsyYLoSH14uKnhNSEuY7AZOE+UuDSsLm/g2+vQNuzIQm0V5H41PLlu1mxIiXCA4O4r77zmL//uJEK4vduw+VJFt79hwud2TKGIiNjSAu7uglPj6Cli2bliRXLVo0LZNstWjRlObNmxAaquRJxFrL4sW7uOOOWcyYsZ5f/jKVRx4Z49nvR25uATt3HmT79oNs336A7dsPHvflq/iSnV1+b6dmzUJp1apZyWeC+4xoSmxsBFu2ZPLRR+vYvv0gAAMGtGbcuK6MG+dOFtm58yA7dpS+HGLXrkNlvvBVVFIQExNOfLx7zfXr95KWlkV0dBMuuKAHl13WhxEjOtbrzx0lYb6jJKw++fEDeHcCXP4dtBnsdTQ+t2pVGsOHv8SuXYcAaNIkmISESBISmtGqlbskJDQr81jx7ZYtm2rERMQHCgoKuf32z/jnP7/nrLOSefPNicTGRvj0NQoLLXv2HGbLlswyl82b3fX27QfYs+fwcbVsoaFBJZ8FpS/x8REkJEQe91jTpqGVxmGtZcmS3Xz44Vo+/HAdc+ZsK7d+rmnTENq2jS753ImPjyi6drfj448+HhcXUSbBys8vZNasjUybtpy3315FZuYRYmObMn58N/r1S6B373h6925F27ZR5Y5uWmvJyMhm48Z9ZGXl0bVrLK1bR3pakqAkzHeUhNUnmRvh2U4w8hnod73X0fjFoUO57Nx5kISESKKiwhpU7ZNIffLii4uZPHk67dpF8/77l9OrV3yVn1tYaNm27QDr1mWweXMmmzfvZ8uWA0WJ1n62bj1wXJlAZGQYHTo0p1275iQlRdG2bTRt20aRlBRdcrtly6Z+/UxIT8/is882kJdXQJs2USWX6OgmPnndI0fymTFjPa+9tpxZszaSlpZVsq158yb06hVP797xhIeHsGlTJhs37mPTpv0cPpxX5jhRUWF06xZL9+5xdO8ey8knu1G8uvq89DIJy8jIYMSIEQDs2rWL4OBg4uPd/8158+ZVeXmgqVOnMnbsWBITy7Z8uuGGG5gzZw65ubls3LiR7t27A3D33Xfz1ltvMXv2bJo3b05wcDBPPfUUp556KldeeWXJ49ZaHnnkEc4666wqxaEkrD6xhfB4NPS9Hs46fpkEERFfmjNnG+efP42srDxeffUixo/vVmZ7RkYWq1als3p1OmvXZrBu3V7Wrcvgxx/3kZNzdIrQGGjTJor27ZvToUMM7dtHl7rdnPbtm9O8uW8SnfokLe0wK1aksWLFnqJrdzs3t4COHVvQsWMMyckxdOwYQ8eOLQgPD2HdugzWrCm+pLNlSybWwk9/2p8pU86tk9rUQBkJu+eee4iMjOS2226r9nOHDh3KE088QUpKSrnb169fz8UXX8zixYtLHrvyyiu5+OKLOf/88/noo4+48847WbRoUZnHP/30U2666SZWrVpVpTgaZJ+wBssEQWwvSF/udSQi0ggMGpTEggWTOe+8aUyY8Bo33jiQ3NwCVq1KZ9WqtDIjOWFhwXTu3IKuXWMZM6YLXbu2pGvXWDp2jKFt22iduFKO+PhmDBvWjGHDkqv8nFGjOpe5n52dx4MPzuaee75i69YDvPXWJQ2yAXJVvPjiizz55JPk5uYyZMgQnnjiCQoLC7nmmmtYvHgx1lomT55MQkICixcv5tJLL6Vp06bVGkErdsYZZ7B+/frjHh88eDDbt/uve5aSMK/F9oGNDadFhYgEtqSkaL755hquu+59Hn98Hi1ahNOzZzwTJnSnZ884evaMp2fPONq3b666TA80bRrK3XcPo2PHFlx//fsMHTqVDz+8ou76vn1xM+xZfOL9qqNVSrVne5YvX84777zDd999R0hICJMnT2batGl07tyZ9PR0li1bBsD+/fuJiYnh8ccfr3Qk7EQ++OAD+vbte9zjM2bM4Pzzz6/RMatCSZjX4vrAiuchKx0i4ryORkQagYiIUF577SKmTBlPZKRqNQPRVVf1JykpmgsvfJ1Bg55j+vTLGTCgjddh1ZnPPvuM+fPnk5rqZvGys7Np164do0ePZs2aNdx0002MGzeOUaNG1ep1brnlFu655x5atWrFM888U+bx3//+92zfvp25c+fW6jUqoyTMa3G93XXGCog409tYRKRRiYpq4nUIUonhwzvy3XfXMXbsK5xxxgu8/vrFx9Xx+VyA1Cdba7n22mu57777jtu2dOlSPv74Y5588kneeustpkyZUuPXefjhh8sd6Sp+/OGHH+a6667zWyKmsWavxRYtJqC6MBEROUavXvHMmXM9PXvGcd5507jzzlm89toyZsxYz9y521i3LoP09KwqrwRSX5x99tm88cYbpKenA+4syi1btpCWloa1lokTJ3LvvfeyaNEiAKKiojh48KDP47j55pvJysri888/9/mxQSNh3otsA01i3EiYiIjIMRITI/nqq6uZNOlt/vKXb8rdJyYmnNWrf0VCQvVWCghUffv25e677+bss8+msLCQ0NBQnn76aYKDg7nuuuuw1mKM4e9//zsA11xzDddff32NC/MrYozhzjvv5MEHHyxppeFLalERCF4b6s6UvOxrryMREZEAZa0lLc0t37Z3b3bJGrqbNu3nT3/6gilTxvOznw2o9nEDpUVFQ6AWFfVRXB9Y+wZY6xrwiIhI1WVuAlsAMZ1PuGt9ZowpWUGgNGstzz67iOnT19UoCRPvKAkLBHF9YOl/4PBONz0pIiIVs4WwawH8+D78+N7Rmtr4FOh+KfS4FJp39C6+/COw/VvY+DFsngnNWsPZT0NMJ7+8nDGG8eO7MXXqD2Rn551weScJHErCAkFs0RmS6SuUhImIlKcgD7Z8DuvfdcnX4Z2ujKPt6XDmQ24WYc3r8O0f3CXxFJeQdbsEotv5P77MTS7p2vgxbJ0FeYchOAzanAa75sFL/d2Zh32u9cuMx7nnduPJJ+cza9ZGxo2r/hmUxTVWUnM1Ke9SEhYI4orOkMxYDskjvY1FRCRQFBbAtq9g9TRY9xbk7IXQZpA8BrqcBx3HQtPYo/sPuMUlQ2vecAnZV7e5y4Bb4cx/+D75ObDZxbb6NUhb4h5r3hF6/RQ6joF2Z0FYJBzYAjOuhk+udwnkqGcgopVPQxk2LJlmzUKZPn1ttZOw8PBwMjIyiI2NVSJWQ25h9gzCw6u3uoEK8wPFvxOg03gY/ZzXkYiIeMcWwvbvXBK19k3I2u0Sr84ToPtlkDwKQqr4h27fOpj3ACyf6kbLUn9b+/gO73ZxrX4NdnznHmt9qhtx6zQOWnQrP9mzhbDoUfjmDxAWDaOehS4Tah9PKRde+Drz5+9gy5abq5VM5eXlsW3bNnJycnwaT2MTHh5OUlISoaFlp4NVmF8fxPZWmwoRaZyshT2LYNVrLvk6tM0lWh3HQY/L3IhXaET1j9uiqxt1yj3gRsSiO0C3i6p/nPwjbgRr+VTY/IlLqOL6wNC/uMSwKrVeJsiN1HUYCR9dCe+dB/0mw1mPVj2pPIFzz+3GO++sZvHiXZx0UusqPy80NJSOHT2sofPAkSP5bNmSSXR0E0/beigJCxRxfWD58+6X26iHrog0AukrYM00N6W3fz0EhbqpxjMecCNfYVG1fw0TBGNegoPb4eMrXd1tm8FVjG85LHsOVr4MORkQmQSn/B/0vOJoGUl1xfWBK+bCd3fD/L9D2lKY8DZEVj1pqsjYsV0xBqZPX1utJKyhSk/PYvnyPaxfv5dNm/azadN+Nm501zt3HsRaePDBs/nd707zLEZNRwaKpVPg05/D9RuhebLX0YhIfZB70NU/LX8e9q9zf9z99flhLRzYBJFtXcF5TeUdhlWvwuKnIG2xS5LaDXcjXl0ugKYtfRZyGVlp8NpgOJIJV8ypuJ3FkQNuqnH5c7BrvksMu5wHfa5zo1hBwb6Lae1b8PFVEN4CznsXEsudsaqWQYOepbDQMm/ez3wQoHcOHcpl/fq9rF2bwbp1Gaxdu5d16zLYuvUAsbFNSUqKJikpmrZto0hKiiYxMZIdOw6ybNkeli93l927D5ccLzjY0K5dc5KTY+jYMYbkZHcZNCiJbt1iK4mk9jQdWR/EllpDUkmYiFTEWtj+jZsaW/Mm5GdByx4uufn8l3DBh74tQM8/4voYLnoUdi+E+P4w7lWI7VW94+xbD0v+7eI+st8dZ/jj0G0iNEvwXbwViYiHCz5yidjbY+Hy744W9Vvr6ruWPeuS2vwsN2I17GHoeSVExPknpm4XQUwXeHcCvH46jH7eJaO1cO653bjzzi/YufMgrVv7YCSxDu3fn8Nzzy3i6acXsn793jLb2raNomvXWIYP78i+fdls23aAefO2k5aWVWa/iIhQeveOZ+zYrvTp04q+fVvRtWssSUnRhIQE3iyTkrBAUdKmYrkr7hQRKS0r3fUTXPE87P/RTdX1nORaHrQ+FX54DL642dVU1fIPOQCHd8GSp90laze07AmD73YjWP8d4Ard+/+i8oTPWtg0A354HDbOcKNIXS+ClBuh7Wl135y6ZTc4/314cwS8dz6MfdUlmMuehb2rITTSvad9r3ctLuoivlb94cr58P7F8OHlkL4MTruvxmUp557bnTvv/IKPPlrHdded7ONg/WPdugwee2wuzz+/mMOH8zjjjA5cc00K3brF0rVrS7p0aUmzZuWPvubk5LNjx0F27DhI69aRdOzYgqCg+nOGp6YjA8l/2kG7YTD2Za8jEZFAkbkJFjzkpsfys91nRJ9roeuF7qzBYoUFbpTnwGa4elXNp/XSl8P8f7gpucI8VxR/8m/cVJwx7uzAmde4flidxsHoqce3WyjIc8ng/L+74zVLhH43uEJ0H9Q+1drq1+HDUolqmyEu8eo20bWU8EJBLnx+Iyx7BjqfBxPeqtHUp7WWDh0e4eSTW/Puuz5Ixv0kKyuP2bO38Pjj85g+fS2hocFcfnkffvObUxtcPZumI+uLuD5HOz+LSOO2ZwnMf9AlMybITYudclvF04BBwTDyGTdK9fXvqt/uZtcCmPsX1ww1tBn0+zmc9Gs3elRaswQ35bn4SXfG4Yt9YcwL0PEcyMtyhewLH3LJYGxvGPOiG5mrTR2Zr/W41E057l0Nva+G2ABYNzE4DEb+B6LawXd3wc650HZItQ9jjOHcc7vxwgtLyMnJJzzc+z/zu3YdYvHiXSxevIslS3azePEu1q7NoLDQEh8fwV13nckNN6SSmNgwFh+vDu//deSo2N6w9Qv3jdaXxZ8iUn/s+B6+v9dN44VGwsk3w4CbISrpxM9t1R9Sb3MjUD2vhPZnnfg5276GOX9xrReaxMCgu+Dkm8o2QT2WMXDSjW5U7sMrXI1VlwtcrVp2uusSP/wJ6DQ2cM/27nON1xEczxgX13d3ufq7GiRh4KYkn3pqAV98sZFzzunq4yArV1BQyNKlu/nmmy18880WZs/ews6dh0q2JyfHkJKSyKWX9uakkxIZPbpLQCSKXmm8P3kgiusDBUdcvcex3z4rkrMfFvzDnR014W1oM8i/MYqIf+z+Ab77E2z4EJrGw2n3Q8ov3Zlz1TH4LtdM9LOfw0+WQGjT4/exFjZ/CnPud4lTRCs4/QFX49UkuuqvFdcHJs2Db++ARY+50bBT/g+ShlYvZjkqsi1EJMCehTU+RHH3/A8+WOv3JCwrK48FC3Ywe3Zx0rWVAweOANChQ3NGjOjEKae0ISUlkX79EoiJ8U1PtIZCSVggKb180YmSsLwsV+w6/++Qs889tmmGkjCR+iZjpesZtfZ/LuEa+jc3ylTT2qTQCDet9b+Rbnpx6P1Ht1kLP34Ac+937Rcik+Csx6DvdTVrhgqu0eiwfx1dv1FqxxjXqmJXzWufw8NDGDmyM9Onr+XJJ8f6dCmibdsO8N13W0suP/ywi/z8QgB69Yrn8sv7cPrp7Tn99A60b9/cZ6/bUCkJCyTFdQnpK1zRbXkKcl3NxZz73AK2Hce6rs3TLzu6dpmI1J0Dm+HbP7oC+hFPQKuUqj1v/wb4/h5Y9QqERLhpwAG3QHhM7WPqcDb0usp9Set+KcT1dj2p5t7vmoM27wgjp0Dvn/quVksJmO8kpLoTH3IP1TgZHz++K+++u5qlS3fTv39ijUPJyMjiiy828dlnG/j8840lrSOaNg1h4MC2/O53QzjttHYMGpREbGwNE/lGTElYIARn8pgAACAASURBVAltBs07lV+cbwtdV+nZf4LMDdB2KIx/HZJOd9tbpbhCThGpG7kH3bqEC//l7odGwSunwMA7YNAfK05uDm53X6KWP+cagQ64FU75ve97UZ35EGz8CD66wtWZ7l0FLbrDOS9Bj8shSB//ASsh1X3m71lc46nd4kW8P/hgbbWSsCNH8vn668189tkGPvtsIz/8sBNrISoqjGHDkvnVr05h6ND29O+fQGhoDWuXjxxwZ/yueAHO+LtvWqrUREEumGBPa7D1Wxho4vq46chi1rpvRN/e4Ua64vu7M5M6nlP2m2d8f3cWVc5+33yTFpHyFRa4GszZd7r+WT2ugNP/5orov/gNzLkX1r/jzhhMKNWnKSvdJW1LnnTH6PdzOPWP/mvZEBEHZz3i1imM6+u+tHW9SCf91AcJA9z17gU1TsISEyMZOLAtH3ywljvvPKPSfXftOsRHH61j+vS1fPLJjxw+nEdoaBBDhrTjz38extlndyI1tU3Nk65i+TnurNq5f3PLQEW1cyd25B5w7UvqirWubvLbO9wIdO+r6u61j6EkLNDE9nbfXgtyXU3At39wZy817+QaC/a4tPyzjYqnQNKXQlLlv3AiUkNbZsGXv3VfiNoMgfPfc41Si419Gbpf4pYge2UgDLzd9dj64Qk3Ypaf5aYJB99dNytj9JwErQe71wrUsxTleJGtXYH+7tr1xBw/vit33fUlu3cfOm6R6lWr0vjf/1Yyffo65s3bDkC7dtFcdVV/xo3rWlTc76Op6sJ8WP4CfP9ntzh7h1Fw+l9dA+APLna/L0cy4ZTfVXyMvKyifnlT3bR/TZuab/0Kvv497JrnBj2qctaxHykJCzRxfdx/2DdHwPZv3VkyI550jQQrq92I7++u9yxREibia/vWwVe/gx/fg+gOblSp28Ty66A6n+vKBb68xRXGz/0rYN3+Q/5c9z2pYjrV7euJbyTUrjgfXKuKu+76kg8/XMe1157Ehg37eP315UybtoKlS3djDJx6ahL3338W48d3o1+/BJ8W8WMtrHvb1UzuW+O+sJzzUtnWKee960Zrv/69S8ROu6/s75UtdHWT3/wBDm13Zw5/MBEu+aLsF6ATSV8B39wOG6a7BHf089DrJ56PDCsJCzTFyVTaUneK+oCby3bFrkiz1u4/Z9pi/8Yn0pjk7HdtHH54DIKbuDMXB9zszgisTHgLNx3ZbaIrJ+hzbdmpSZETSUx1Sf+RA9VrG1JK//4JJCVF8/DDc5gyZSFz57oRryFD2vHYY2O4+OJe/ltf8sBW+PwXruVKbC+XbHWecPwXl+AwGPcahEW7Ly25B9w0uglyo1Zf3ep6piWeAuOmQYuuMO00eHucW//zRJ0EDu1wZx8vn+pKBob+zY1Ol9e6xQNKwgJNXG+Y+LlLxiprlngsY9xzdIakSO0V5rv1BGf/CbIzXBI19H63/E51dBqntWClZorrwvYsck1xa8AYw0UX9eTRR+dy8smtefDBs7nkkt506ODHumFbCEv+A9/8n6t9HPYvOOmmykecgoJh1DPQpLmbts/OcEt0rX/HtVEZ+193MknxlPqFM+C1IfDWaLji+/J/L611nQS++q2rRTvpJleD6a/F2GtISVggaj+8Zs9rleJ6hxXkQXCob2MSqciBrTD7j9DtEug83utoam/bN/D5L91ZyklnwLBHIOEkr6OSxqY4Cdu1oMZJGMADD5zNrbcOpl27OujZtXctfHK9awDcfoRrg1LV6XBj4Mx/ukTsu7vdqNVp97u2Lcf2sGvRBS78EF4f5lZruOTLsqOFB7bCpz+DTTPdezfyGfecAKQkrCGJ7+867u9bc7Txq4i/WOtqNWbd6Go5cvbV7yQs/4hbLmb+P1zd17n/c/361P9KvBDRCqLa17o4Pzw8xP8JWGE+zP+n63sX0tQt6t776ur/7hjjVnxoNwxadKt85DnxFJjwP3jnXHj/IpeUBYW6accvf+tiGv4EpPwioE9KURLWkBSfIZm2REmY+Fd2Bnz2C3ead5shEBblzjaytn4mLenLXXFw2hJ3qvyZD9W8Y72IrySmunqoQJa5yf3u7JjtWqAMf7z2bVeqenJZx3Ng1LMw8xoXQ+5Bt3JM0pkuEawHJ6UoCWtIWnR3xcN7FrtT00X8YeMMmHmtW6h56N/caeVLp7ih/4Nb3ChSfWELYdGj7syrJs3h/Pfd2Y0igSAh1Z1dmLOv+muI1oXV01x7Cayr2/Li706fq+HwDncGZkiESwJTfhnQo1+lKQlrSIJDXZ8xFeeLP+Rlw1e3wZKn3P+zCz86OvraeqC73jmv/iRhB7bCzKtd76/OE1xhcEQrr6MSOSoh1V3vXgQdRngbS2m5B2HWTa7jfetBMPYVb0edBv7BTV+2OrlejH6VpiSsoWmV4hbora/TQhKY9v8I71/sWqAMuNWdKVi6TUNcX3eq+a550H2id3FW1eppbjq1MM8V7fa9Tr8vEnhKd84PlCRs13zX5T5zAwz6k6vh8noJLGOg28XexlBDSsIamvj+rjDx8C7/LYcijcv692DGT93w/gXTy2+5EBwGrU5yH9CB7EgmfH4jrPqv+wZ/zssBe9aUCE1butVSalmc7xO20J20MvtO15fyki+Prl0sNaYkrKEpKc5frCRMaqcwH769E+b/3X0jP/d/lS+1kzjQfQEoLPC8C3W5tn0NH/3Edd0efI9bZNvrb/AiJ5KQ6kaYvXRoJ3x8FWz5zI04jZwSmDVq9VD9qFyTqovr5673qC5MauHwLnjzbJeA9b8BLvv2xGsdJg6EvMOwd1WdhFgi/wh8cYvrF/T9vbD5M9dlvFhBriu8f32Yq5u87FsYcrcSMKkfElPhwCa3ALwXNnwIL/VzZz+OfAbGv6EEzIf0KdTQhMdAdLKWL5Ka2/YNTL/ETd2d85JbX60qEksV59dVi5SsPfDehe4PRIvu7sxNLGBcDG2GuCnSPYugz3VuORS1npD6pKRz/kJIHl13r5t/xHW9X/QoxPdzSwbV9bqnjYCSsIZIyxdJsYxVsPhJt2THidZYsxYWPuwW0o3pBBd9AvF9q/5aLbq4Ng+75kHfa2sXd1WkLYN3z3WJ2Pg33AkBRzJdErjjO9j5PayZ5ho4Tngbul7g/5hEfK1V0ZqjuxbUXRKWsRo+vNx9mT/pJjjj7ydeL1VqRElYQ9QqBX58300NVWXxb2mY1r0NH/8U8g7B8ufg9AfgpF+X3z8n9yDMvM41X+1ygVt8urqLBpsgSDilborzf5zu/kg0iYZLv3ZTNuCSwOSR7gKumNjawKxRE6mK8Bi3aHVdFeevf9/9boU0Vd+8OqCasIYoPgWwrgu4ND6FBa5x4fsXuQXhJ81367h9cTO8MRwyN5bdP2MlvDIQ1r0FZzwIE96qfgJWrPVASF/qeor5g7Ww4CF4dwK07A5XzDuagJXHBCkBk/ovIdWNhPnbkqfh/QsgthdctUQJWB1QEtYQtervrjUl2fjk7IN3xsPcv0Lfn8ElX7kk5fwP3DIeexbBi31dh3trYfXrLgHL2QsTP3fd72vTLytxoDur0h81iQW5bnHgr26Dbhe5EbCotr5/HZFAk5AKh7bB4d3+Ob617kzoz34ByWPg0i/1u1VHNB3ZEEUnQ1i0W75IGo+0pfDeBXBwK4z8j1sDsZgx0OcaNyI28zq31MiixyBjhSteH/+Gbz50i4vzd82DNoNrf7xi2RluZG/bV65B5JB76s2yJCK1Vjzau3shdBrr22MX5MGnP4MVL0Lf6+Hsf+vM4TqkT7GGyBgV5zc2q1+HVwdDfjZc+lXZBKy06PZw8Scw4in3zfqkm+CSL3z3rTeyNUS2dcXxvrJ3Dbw6yBXaj/0vnHavEjBpXFqdBJjK68Ksrf5xcw+6kfMVL8KQP7v+X0rA6pQ+yRqqVikuCbOFXkci/rbocfjwMvdBfeXCE49AGQMpv4Bf7YPhj7pu976UOBB2+6g4f/PnLgE7kgkTv9DC9NI4hUVByx4V14XN/Ss8FQebZlb9mId3w+tnwpbPXanC4Lu0dJcHlIQ1VPH93dmR+zd4HYn40/x/whc3uTMaL5lVvVUS/PWBmzgQ9q2D7L21O87SZ+DtMW5kbdI8aDvEN/GJ1EeJqcePhFkLs+9yJ+IU5rsTVn6cfuJjHdoJbwxzo8wXfOBKFcQTSsIaqtLLF5VnyxfwQh/VjdVnc+6Hr38H3S+D8a/7fkSrploX1YXV9JT6wgL48lb4dDK0Pxsu/+7E3fpFGrqEAXB4Jxza4e5bC9/eAXPugz7XwnU/uhVT3r8Q1r1b8XEObncJ2MFtcNEM6HhOnYQv5VMS1lDF9gYTXH5d2Lq33QhDxgq3OLPUL9bC7D+5S6+rXJ1UcKjXUR2VMAAwNVvvriAPpl8KC//leppd8EHN22WINCQJRcX5uxa4z4CvboV5D7hlxUY9AxFxMPEz9/s3fSKsefP4YxzYCm+c6ZK5i2ZqAe4AoAq8hiok3NUQHDvStew5N8KQeCpk7a67BoDiG9bC1/8HC/7hzmQa+Z/AK1Jv0tz18KpucX5hPnx0petXNuxfMOAW/8QnUh+1SnG/67vmweZPjq6EcdYjR0sLmjR3ydU741ydaGEe9LzCbTuwGd44y51pfPGn0PpU734WKRFgn97iU8XF+cXmPej6LHUYBRM/hbanuVOepX6w1jVcXfAPSPlVYCZgxRIHuj8WVT1jyxbCzGth7Rtw5j+VgIkcK7SZa6K68CGXgKXeVjYBK9YkGi78GJLOcF9qVrwImZvcAvY5+9xomRKwgBGgn+DiE/H9Xc+o7Az46nduMdYel8P577lf6ITUsjUGErisdf+GPzwGA34Lwx8P3AQMXBKWtdvVnZyILXR9y1a+DKfdD6m3+j8+kfoo4RTIz4FT/+hWt6jo5JqwSLjgQ9cXcMY1riFzbqZLwBJPqduYpVIB/CkutRZfVJz/zjhY8E9IubGofqiogLt0jYEEtnkPuG/AKTe6kaJAP5W8dammrZWxFmb9BpY9C4PuhEF/9H9sIvXVaffCee/B0PtP/BkQGuFqKoubu06cVVSvKYFENWENWfHyRTvnwuB7ju8DU1xjsHsBdJngSYhSBUv+486C6jnJ9fUK9AQM3FlawWEuCet2Ufn7FI/uLX7CTa0MubduYxSpb6KS3KWqQsLdkmUFR9xtCThKwhqyiFZu5CS+P/S7/vjtoRGuxkB1YYFrzRtuPbdO42D084E9BVlaSBP3/66ykbDv7jo6ulfZ1IqI1JwxSsACmJKwhm7E45VvT0iFjR+5UQn9EQwsm2a6wtq2Q93ajoHUhqIqEgfCypdc36+g4LLbFv/b9Tnrc139Gd0TEfGxevK1WvwmIRWy9lStgFrqzo7v4b0LXb+3Cz5wo5b1TeJAtzbdvjVlH/9xOsy6ETqNh5FP15/RPRERH9OnX2OXWFScr35hgSN9uTuZIrKN62jdpLnXEdVMYlFxful+YbsWuGasrU6C8dO0WLCINGpKwhq7uH6us77qwgJD5kb43ygIaeoaKjZL8DqimmvZDcKij9aFZW6Cd8ZDRDxcMN21SRERacT0NbSxC20KcX00EhYIsva4BKwgBy79pv6vl2iC3EjrrnmuSeTbY91ZWpfMgmaJXkcnIuI5jYSJqwsrXo9MvHHkALw1Bg5td00W43p7HZFvJA6EtKXw3vmwfz2c9647I1dERPybhBljxhhj1hhj1htjbi9ne3tjzBfGmB+MMUuNMWP9GY9UIDEVcjLc2mJS9/Jz4L3zIH0ZTHgL2gz2OiLfSRzo1q/b9jWMeQHanel1RCIiAcNvSZgxJhh4EjgH6AVcbow59ivwncAb1tqTgMuAp/wVj1SiuIuy6sLqXmE+fHgFbP3SJSkdz/E6It9qM8TVfg3929GFhEVEBPBvTdhAYL21dgOAMWYacB6wstQ+Foguut0c0CKGXojrB0Ghri6sou7m4nvWwqc3wPp34KxHXUf8hqZZAvxqX/3rcSYiUgf8OR3ZFtha6v62osdKuwe40hizDfgI+HV5BzLGTDbGLDDGLEhLS/NHrI1bSBOI66s1JOvat3fA8ufcYrwn3+R1NP6jBExEpFxeF+ZfDrxgrU0CxgIvG3N850Zr7RRrbaq1NjU+Pr7Og2wUElNhz0IV59eV5c+7Rbn7/RxOu8/raERExAP+TMK2A+1K3U8qeqy064A3AKy13wPhQJwfY5KKJKS6NgKZG72OpOFLXw6f/wraD4cRT2rJHhGRRsqfSdh8oKsxpqMxJgxXeP/+MftsAUYAGGN64pIwzTd6oaQ4X1OSfpV7CD6Y6JqYjn3l+DUVRUSk0fBbEmatzQduBGYCq3BnQa4wxtxrjJlQtNutwM+MMUuA14CrrdV8mCfi+kBwmOrC/Mla+PyXsHcNjHtVDUtFRBo5v3bMt9Z+hCu4L/3YXaVurwRO82cMUkXBYRDf39WFiX+seAFWvgyD73FTkSIi0qh5XZgvgSQh1fUKs4VeR9LwlK4DG3Sn19GIiEgAUBImRyUMgCOZsP9HryNpWFQHJiIi5VASJkclpLpr1YX5jurARESkAkrC5KjYXhASruWLfGnly0V1YHerDkxERMpQEiZHBYdCfIraVPjKga0w69fQ9nTVgYmIyHGUhElZCQNUnO8L1sIn14MtcAtzqw5MRESOoSRMykpIhbxDsHet15HUb0unwOZP4Ix/QEwnr6MREZEApCRMykosKs5Xv7Cay9wIX90K7UdA/597HY2IiAQoJWFSVsseEBIBO+d5HUn9ZAthxjVggmD0VHctIiJSDr92zJd6KCgEWg+End97HUn99MMTsO0rGPUsRLf3OhoREQlg+poux2s9GPb8AHnZXkdSv+xdC9/cDh3HQp9rvY5GREQCnJIwOV6bIVCYr1YV1VFYADOvgeAmMHIKGON1RCIiEuCUhMnxWg9y1zu+8zaO+mThw+79Gv4YRLX1OhoREakHlITJ8SLioEU32KG6sCrJWAWz74TO50HPK72ORkRE6gklYVK+NoPdyI61XkcS2IqnIUObwcinNQ0pIiJVpiRMytdmCGSnwf4fvY4ksC14CHbOheGPa3FuERGpFiVhUr7Wg921WlVULGMVfHcXdLkAelzudTQiIlLPKAmT8sX2grBoFedXpDAfZlwNoZFw9r81DSkiItWmZq1SvqBgaH2qivMrsuAh2DUPxr0GzRK8jkZEROohjYRJxdoMgfRlkHvQ60gCS8ZKNw3Z9ULofqnX0YiISD2lJEwq1mawWwtR60geVTINGQUjntI0pIiI1JiSMKlY4qmAUV1YafP/Cbvmw4gnNQ0pIiK1oiRMKhYe4wr0dYakk7ESvr8bul0M3S/xOhoREannlIRJ5doMccX5ttDrSLz3xS2uKeuIJzUNKSIitaYkTCrXZjAc2Q9713gdibc2fQKbP4FBf4KIVl5HIyIiDYCSMKlcmyHuujHXhdlC+Pr3EJ0M/X/pdTQiItJAKAmTyrXoBuEtG16/sO2z4YNLITvjxPuuegXSlsDQv0JIE//HJiIijYKatUrljDm6mHdDUVgAn/4cMlbAwS0w8XMIjSh/3/wc+PaPkDAAeqgnmIiI+I5GwuTEWg+GvasgZ5/XkfjG6lddAtbnOrf49vRLXf+v8vzwOBzcCmf8A4x+XURExHf0V0VOrLgubOccb+PwhYJc+O5uaHUSjJriznTcMN2NjFlbdt/sDJj7F+g4Ftqf5U28IiLSYCkJkxNLPMWNAjWEurClz0DmRlffZYIg5Rcw6E5YPtUtRVTa3L+6JZvO+Ls3sYqISIOmmjA5sbBIiO9f/+vC8g7DnPsg6QxIHn308SH3wqGdMOd+aNYaUn7pErXFT0DvqyGuj2chi4hIw6UkTKqm9WBY+ZIrag8K9jqamln0GGTthglvl222agyMfNpt+/xGaJYIa98CEwxD/uxdvCIi0qBpOlKqpu0QyDsE6cu9jqRmcvbB/Aeh03j3sxwrKATGvw6tB8GHl7vi/QG3QFRS3ccqIiKNgpIwqZrWg911fZ2SnP8gHMmEoX+peJ/QCLjgA2jeyXXFP+X3dRefiIg0OkrCpGqad4SIhPqZhB3aCYsehR6XQ3y/yvdtGguT5sNVS6BJ87qJT0REGiXVhEnVGOMK2rfOcq0c6tMC1nPuh8I8OO3equ0fFukuIiIifqSRMKm65NFwaAdkrPQ6kqrb/yMsmwJ9r4eYzl5HIyIiUkJJmFRdh5HuetNMb+Ooju/vhaBQGPQnryMREREpQ0mYVF10e2jZAzZ/4nUkVZO5yS2+3W8yRLbxOhoREZEylIRJ9XQYBdu+cgtbB7oF/3Rd8VNv8zoSERGR4ygJk+pJHu0SsO3feh1J5Q7vhuXPQa+r1OtLREQCkpIwqZ52Z7oaq9rWhRXkwrq3IWuPb+I61qJHIP+Ien2JiEjAUhIm1RPaDNoOrXldWGE+LJsKU7vD+xe5wnlfy9kPi5+EbhOhZTffH19ERMQHlIRJ9SWPhrSlrglqVRUWwMr/wvM94ZProGmca5y6ZZbv41vyFOQehIG3+/7YIiIiPqIkTKqvwyh3veWzE+9rC2HNG/BiX/j4J24k7bz3YNI86PkT2LuqesncieRlwcJHoOM5kHCS744rIiLiY0rCpPpa9Yem8VWrC/vgEph+qeuwf+6b8JNF0GWCu99+uNtn6xe+i23Zc5CdBgP/4LtjioiI+IGSMKk+E+Qat27+1I10VWTPYlj3liuOv2opdLvYPbdYfH9oEuO7KcmCXFjwD1ezlnS6b44pIiLiJ0rCpGaSR7szG9OWVrzPwofd9OPAP0BQ8PHbg4Kh3TDfjYStehUOboVT7/DN8URERPxISZjUzImWMDq0E1a/Bn2uhfCYio/TbjhkbnDd7WujsADmPQDxKZA8pnbHEhERqQNKwqRmIltDXN+KW1UsftK1ozj5N5Ufx1d1YevfhX1r4NQ/uHozERGRABfidQBSj3UYBYsfh7zDbtqxWF4WLPk3dDkfYjpXfozYXhDRytWF9bnmxK95JBP2rnEJ197VRy/71kGLrtD1otr9TCIiInVESZjUXPJoWPgQbPvatYQotvIlyNkLA2458TGMcVOSW2eBtZWPYs35C8y+8+j9oBCI6eIWFe90LvS+uvzaMxERkQCkJExqru1QCAl3dWHFSZgtdH26ElLd9qpofxasmQb71kLL7uXvc+SAq/lqPwJO+rVLvJp3guBQ3/wsIiIidUxJmNRcaFNoewZsKlUXtvFjN1U49tWq12a1K1UXVlEStvIlyDsEp/8NEk+pXdwiIiIBQIX5UjvJo13X+wNb3f2F/4LIJNcTrKpiOkNUu4r7hdlC+OEJaH2qEjAREWkwlIRJ7SQXLWG0+VPXnHXLLDddWJ1pwuLu+Vu/KL/56+bP3ehayo2+iVlERCQAKAmT2ontDc1au7qwRY+4syT7/az6x2k3HLLTIX358dt+eNydQdltYu3jFRERCRBKwqR2jHGjYZtmuI71va+B8BbVP067s9z1sVOSmRthw3ToNxlCmtQ+XhERkQChJExqr8NoyD1QteasFYlu5/p8HZuELX7KrTfZ74baxykiIhJAlIRJ7XU4GzDQ5Txo0aXmx2l3Fmz7yiVz4Jq+Ln8Oul4IUW19EqqIiEigUBImtRcRD+e/D8OfqN1x2g13I2p7fnD3V70KOftcob+IiEgDoyRMfKPz+NqPVrUb5q63FHXPX/wExPeretNXERGResSvSZgxZowxZo0xZr0x5vYK9rnEGLPSGLPCGPOqP+ORANcsAeL6uCRs+7eQtgRSfq0FuUVEpEHyW8d8Y0ww8CQwEtgGzDfGvG+tXVlqn67AH4DTrLX7jDGt/BWP1BPthsOyZ2DBQ+4sy55XeB2RiIiIX/hzJGwgsN5au8FamwtMA847Zp+fAU9aa/cBWGv3+DEeqQ/aD4f8bPjxPehzHYRGeB2RiIiIX/gzCWsLbC11f1vRY6V1A7oZY2YbY+YYY8aUdyBjzGRjzAJjzIK0tDQ/hSsBIekMwLhL/194HY2IiIjfeL2AdwjQFRgGJAFfG2P6Wmv3l97JWjsFmAKQmppq6zpIqUPhLVyBftN4iOnkdTQiIiJ+488kbDvQrtT9pKLHStsGzLXW5gEbjTFrcUnZfD/GJYHu4k+8jkBERMTv/DkdOR/oaozpaIwJAy4D3j9mn3dxo2AYY+Jw05Mb/BiT1AdBIe4iIiLSgPktCbPW5gM3AjOBVcAb1toVxph7jTETinabCWQYY1YCXwC/s9Zm+CsmERERkUBhrK1fJVapqal2wYIFXochIiIickLGmIXW2tTytqljvoiIiIgHlISJiIiIeEBJmIiIiIgHlISJiIiIeEBJmIiIiIgHlISJiIiIeEBJmIiIiIgHlISJiIiIeEBJmIiIiIgHlISJiIiIeEBJmIiIiIgHlISJiIiIeEBJmIiIiIgHlISJiIiIeKDaSZgxJsgYM8kfwYiIiIg0FhUmYcaYaGPMH4wxTxhjRhnn18AG4JK6C1FERESk4QmpZNvLwD7ge+B64A7AAOdbaxfXQWwiIiIiDVZlSVgna21fAGPMs8BOoL21NqdOIhMRERFpwCqrCcsrvmGtLQC2KQETERER8Y3KRsL6G2MO4KYgAZqWum+ttdF+j05ERESkgaowCbPWBtdlICIiIiKNSYVJmDEmHLgB6AIsBaZaa/PrKjARERGRhqyymrAXgVRgGTAWeKhOIhIRERFpBCqrCetV6uzI54B5dROSiIiISMNX1bMjNQ0pIiIi4kOVjYSlFJ0NCe6MSJ0dKSIiIuIjlSVhS6y1J9VZJCIiIiKNSGXTkbbOohARERFpZCobCWtljPltRRuttf/yQzwiIiIijUJlSVgwEMnRjvkiIiIi4iOVJWE7rbX31lkkIiIiIo1IZTVhGgETERER8ZPKkrARdRaFiIiI7zybYQAAEA5JREFUSCNTYRJmrd1bl4GIiIiINCaVjYSJiIiIiJ8oCRMRERHxgJIwEREREQ8oCRMRERHxgJIwEREREQ8oCRMRERHxgJIwEREREQ8oCRMRERHxgJIwEREREQ8oCRMRERHxgJIwEREREQ8oCRMRERHxgJIwEREREQ8oCRMRERHxgJIwEREREQ8oCRMRERHxgJIwEREREQ8oCRMRERHxgJIwEREREQ8oCRMRERHxgJIwEREREQ8oCRMRERHxgJIwEREREQ8oCRMRERHxgJIwEREREQ8oCRMRERHxgJIwEREREQ8oCRMRERHxgJIwEREREQ8oCRMRERHxgF+TMGPMGGPMGmPMemPM7ZXsd5ExxhpjUv0Zj4iIiEig8FsSZowJBp4EzgF6AZcbY3qVs18U8Btgrr9iEREREQk0/hwJGwist9ZusNbmAtOA88rZ7z7g70COH2MRERERCSj+TMLaAltL3d9W9FgJY8zJQDtr7YeVHcgYM9kYs8AYsyAtLc33kYqIiIjUMc8K840xQcC/gFtPtK+1doq1NtVamxofH+//4ERERET8zJ9J2HagXan7SUWPFYsC+gBfGmM2AYOA91WcLyIiIo2BP5Ow+UBXY0xHY0wYcBnwfvFGa22mtTbOWptsrU0G5gATrLUL/BiTiIiISEDwWxJmrc0HbgRmAquAN6y1K4wx9xpjJvjrdUVERETqgxB/Htxa+xHw0TGP3VXBvsP8GYuIiIhIIFHHfBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMBEREREPKAkTERER8YCSMPn/9u4/1q+6vuP480ULMjCK/BjRwvghndoxEVY6Fh0yIQqRADGIMAgE2RiLKLqxBafRjcRsyBR1MhRBisbAEN1oprMS7MSZCK3AgFIZtTB+pEK3Ij90K3S898c5N357Wyy33G8/337v85E093s+53PP93PuJ5/b1/2c8z0fSZLUgCFMkiSpAUOYJElSA4YwSZKkBgxhkiRJDRjCJEmSGjCESZIkNWAIkyRJasAQJkmS1IAhTJIkqQFDmCRJUgOGMEmSpAYMYZIkSQ0YwiRJkhowhEmSJDVgCJMkSWrAECZJktTAUENYkqOT3JtkZZILNrH/T5Lck+TOJDcl2WeY7ZEkSRoVQwthSWYBlwLHAPOAU5LMm1TtdmB+Vb0euB74+LDaI0mSNEqGORO2AFhZVauq6hngWuD4wQpVtaSqft5v/gDYa4jtkSRJGhnDDGFzgIcGth/uy57PWcC/bGpHkrOTLEuybM2aNdPYREmSpDZG4sb8JKcB84GLN7W/qi6vqvlVNX+PPfbYuo2TJEkagtlDPPYjwN4D23v1ZRtIchTwIeDNVbVuiO2RJEkaGcOcCVsKzE2yX5IdgJOBRYMVkhwMfB44rqoeG2JbJEmSRsrQQlhVrQfOBRYDK4Drqmp5kguTHNdXuxh4KfDVJHckWfQ8h5MkSRorw7wcSVV9E/jmpLKPDLw+apjvL0mSNKpG4sZ8SZKkmcYQJkmS1IAhTJIkqQFDmCRJUgOGMEmSpAYMYZIkSQ0YwiRJkhowhEmSJDVgCJMkSWrAECZJktSAIUySJKkBQ5gkSVIDhjBJkqQGDGGSJEkNGMIkSZIaMIRJkiQ1YAiTJElqwBAmSZLUgCFMkiSpAUOYJElSA4YwSZKkBgxhkiRJDRjCJEmSGjCESZIkNWAIkyRJasAQJkmS1IAhTJIkqQFDmCRJUgOGMEmSpAYMYZIkSQ0YwiRJkhowhEmSJDVgCJMkSWrAECZJktSAIUySJKkBQ5gkSVIDhjBJkqQGDGGSJEkNGMIkSZIaMIRJkiQ1YAiTJElqwBAmSZLUgCFMkiSpAUOYJElSA4YwSZKkBgxhkiRJDRjCJEmSGjCESZIkNWAIkyRJasAQJkmS1IAhTJIkqQFDmCRJUgOGMEmSpAYMYZIkSQ0YwiRJkhowhEmSJDVgCJMkSWrAECZJktSAIUySJKkBQ5gkSVIDhjBJkqQGDGGSJEkNGMIkSZIaMIRJkiQ1MNQQluToJPcmWZnkgk3sf0mSf+j335Jk32G2R5IkaVQMLYQlmQVcChwDzANOSTJvUrWzgMer6gDgEuCiYbVHkiRplAxzJmwBsLKqVlXVM8C1wPGT6hwPXN2/vh44MkmG2CZJkqSRMHuIx54DPDSw/TDw289Xp6rWJ3kC2A34r8FKSc4Gzu43n05y7zS3dffJ76mxZn/PLPb3zGFfzyzbSn/v83w7hhnCpk1VXQ5cPqzjJ1lWVfOHdXyNFvt7ZrG/Zw77emYZh/4e5uXIR4C9B7b36ss2WSfJbODlwH8PsU2SJEkjYZghbCkwN8l+SXYATgYWTaqzCDijf30i8J2qqiG2SZIkaSQM7XJkf4/XucBiYBbwxapanuRCYFlVLQKuBL6cZCWwli6otTC0S50aSfb3zGJ/zxz29cyyzfd3nHiSJEna+nxiviRJUgOGMEmSpAbGPoQl2TvJkiT3JFme5Ly+fNckNya5r//6ir48ST7TL6V0Z5JD2p6BtkSSWUluT/LP/fZ+/dJYK/ulsnboy106axuXZJck1yf5UZIVSX7H8T2+knyg/11+d5Jrkuzo+B4fSb6Y5LEkdw+UTXk8Jzmjr39fkjM29V6jYOxDGLAe+NOqmgccBrynXz7pAuCmqpoL3NRvQ7fM0tz+39nAZVu/yZoG5wErBrYvAi7pl8h6nG7JLHDprHHwaeBbVfVa4CC6fnd8j6Ekc4D3AfOr6kC6D32djON7nCwEjp5UNqXxnGRX4KN0D4hfAHx0IriNmrEPYVW1uqpu618/RfcLeg4bLpl0NXBC//p44EvV+QGwS5JXbuVm60VIshfwduCKfjvAW+iWxoKN+9uls7ZRSV4OHE73SWuq6pmq+imO73E2G/iV/tmSOwGrcXyPjaq6me5pCYOmOp7fBtxYVWur6nHgRjYOdiNh7EPYoH4q+mDgFmDPqlrd7/oJsGf/elPLLc3ZSk3U9PgU8OfAc/32bsBPq2p9vz3YpxssnQVMLJ2lbcN+wBrgqv7y8xVJdsbxPZaq6hHgb4EH6cLXE8APcXyPu6mO521mnM+YEJbkpcDXgPdX1ZOD+/oHxPqsjjGQ5Fjgsar6Yeu2aKuYDRwCXFZVBwM/4xeXKgDH9zjpLykdTxe+XwXszIjOcGg4xm08z4gQlmR7ugD2lar6el/86MRliP7rY335C1luSaPrjcBxSR4ArqW7TPFpumnqiYcTD/apS2dt2x4GHq6qW/rt6+lCmeN7PB0F3F9Va6rqWeDrdGPe8T3epjqet5lxPvYhrL/+fyWwoqo+ObBrcMmkM4AbBspP7z91cRjwxMA0qEZcVX2wqvaqqn3pbtj9TlWdCiyhWxoLNu5vl87aRlXVT4CHkrymLzoSuAfH97h6EDgsyU797/aJ/nZ8j7epjufFwFuTvKKfPX1rXzZyxv6J+UneBHwPuItf3CP0F3T3hV0H/Brwn8BJVbW2H9ifpZvi/jlwZlUt2+oN14uW5Ajg/Ko6Nsn+dDNjuwK3A6dV1bokOwJfprtXcC1wclWtatVmTV2SN9B9CGMHYBVwJt0fmI7vMZTkr4B30X3y/XbgD+ju93F8j4Ek1wBHALsDj9J9yvGfmOJ4TvJuuv/rAT5WVVdtzfN4ocY+hEmSJI2isb8cKUmSNIoMYZIkSQ0YwiRJkhowhEmSJDVgCJMkSWrAECZpWiSpJJ8Y2D4/yV9O07EXJjlx8zVf9Pu8M8mKJEuG/V6baccDSXZv2QZJw2cIkzRd1gHvGLXwMPAk9RfiLOAPq+r3htUeSZpgCJM0XdYDlwMfmLxj8kxWkqf7r0ck+W6SG5KsSvI3SU5NcmuSu5K8euAwRyVZluQ/+jVCSTIrycVJlia5M8kfDRz3e0kW0T1RfXJ7TumPf3eSi/qyjwBvAq5McvGk+q9McnOSO/rv+d2+/LK+Tcv7h4hO1H8gyV/39ZclOSTJ4iQ/TnLOQBtvTvKNJPcm+VySjX4nJzmt/3nckeTz/TnP6n+md/fnsdHPXNLom8pfiJK0OZcCdyb5+BS+5yDgdXRPNF8FXFFVC5KcB7wXeH9fb19gAfBqYEmSA4DT6ZYqOTTJS4DvJ/l2X/8Q4MCqun/wzZK8CrgI+C3gceDbSU6oqguTvIVulYXJT9H/fWBxVX0sySxgp778Q/2Tu2cBNyV5fVXd2e97sKrekOQSYCHdGoc7AncDn+vrLADm0T0F/FvAO+jWv5xo6+vong7/xqp6NsnfA6cCy4E5VXVgX2+XF/KDljRanAmTNG2q6kngS8D7pvBtS6tqdVWtA34MTISou+iC14Trquq5qrqPLqy9lm5NuNOT3EG3FNluwNy+/q2TA1jvUOBf+0Wg1wNfAQ7fXBuBM/t73H6zqp7qy09KchvdUjm/QReoJiwaOI9bquqpqloDrBsITbdW1aqq+j/gGrqZuEFH0oXFpf05Hgns35///kn+LsnRwJObab+kEeRMmKTp9ingNmBwrbb19H/09ZfcdhjYt27g9XMD28+x4e+oyWusFRDgvVW1weK8/bqhP9uy5m+sqm5OcjjwdmBhkk/SrUl7PnBoVT2eZCHdTNeEwfOYfI4T57WpcxoU4Oqq+uDkNiU5CHgbcA5wEvDuqZ6XpLacCZM0rapqLd1iu2cNFD9AN6MDcByw/RYc+p1JtuvvE9sfuBdYDPxxku0Bkvx6kp03c5xbgTcn2b2/jHgK8N1f9g1J9gEeraov0C0WfgjwMrqg90SSPYFjtuCcFiTZrw+m7wL+bdL+m4ATk/xq345dk+zTf/hhu6r6GvDhvj2StjHOhEkahk8A5w5sfwG4Icm/0937tCWzVA/SBaiXAedU1f8muYLukuVtSQKsAU74ZQepqtVJLgCW0M00faOqbtjMex8B/FmSZ4GngdOr6v4ktwM/Ah4Cvr8F57QU+CxwQN+ef5zU1nuSfJjuvrXtgGeB9wD/A1w1cCP/RjNlkkZfqibPfkuShq2/ZHp+VR3bui2S2vBypCRJUgPOhEmSJDXgTJgkSVIDhjBJkqQGDGGSJEkNGMIkSZIaMIRJkiQ18P+S3pLezrNAagAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "trn_score_tpr = exponential_moving_average(score_trn_list[:,0], 0.1)\n",
    "tst_score_tpr = exponential_moving_average(score_tst_list[:,0], 0.1)\n",
    "\n",
    "plt.title(\"TPR per samples\")\n",
    "plt.xlabel('Number of samples')\n",
    "plt.ylabel('TPR')\n",
    "plt.plot(train_size, trn_score_tpr, label='Train TPR',color='navy')\n",
    "plt.plot(train_size, tst_score_tpr, label='Test TPR',color=\"darkorange\")\n",
    "plt.legend(loc=\"best\")\n",
    "plt.ylim(0,1.05)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Specificity SPC = TN / (FP + TN):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmEAAAG5CAYAAADGcOOUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nO3de5hddX3v8fd375lJAglEEwRNwICESkANIQVvp1JF5eIB9ajAQaGopfqAitVa21oFfGxL1YpVTjmAiHoU5GhVFAERQa0egQQjAiESASEYIAQItyRz+54/1prJzmQm953fZPJ+Pc969rr81lrftfbszCe/tWavyEwkSZK0bTVKFyBJkrQjMoRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSRplIuLeiDi8dB2S2ssQJmmLRcQrI+KXEbEiIh6NiF9ExJ/Wy/4iIvoi4qmIeCIiFkTEG1rW3SUizo2I++o2v6+np5Y7IklqP0OYpC0SEbsAPwC+ADwbmAacBaxuafb/MnMiMBn4EnB5RDwrIrqA64ADgCOAXYCXAcuBQ9pcd0c7ty9JG2IIk7Sl9gPIzEszsy8zV2bmjzLz1qENM7MfuBiYALwAOAnYC3hTZt6Rmf2Z+XBmfjIzfzjcziIiI+L9EXF3RDwSEZ+OiEbL8ndGxMKIeCwiromI5w9Z97SIuAu4a5htj4+I/xMRyyPi8Yi4OSJ2r5edUm/3yXrff9Wy3mERsSQiPhIRD0fE0oh4Y0QcFRG/q3sH/76l/ZkR8a2I+Ga9vVsi4iUjHG8jIj5a9xAuj4jLI+LZG6pX0uhnCJO0pX4H9EXEVyLiyIh41kgN696ndwNPUYWgw4GrM/OpTdznm4C5wBzgWOCd9faPBf4eeDOwG/Bz4NIh674ROBSYNcx2TwZ2BfYEpgDvAVbWyx4G3kDVW3cK8LmImNOy7h7AeKqewI8DFwJvBw4G/hvwjxGxd0v7Y4H/S9V7+A3guxHROUxN76trfhXwPOAx4LyNqFfSKGcIk7RFMvMJ4JVAUgWPZRFxxZAemZdGxOPAg8AJVD1fK6iCw9LN2O05mfloZt4HnFtvE6oQ8s+ZuTAze4F/Ama39obVyx/NzOHCSk9d0751r978+vjIzCsz8/dZ+SnwI6pw1brupzKzB7gMmAp8PjOfzMzbgTuA1t6u+Zn5rbr9v1EFuJcOU9N7gH/IzCWZuRo4E3hLHWhHrFfS6GcIk7TF6tDzF5k5HTiQqsfm3JYmv8rMyZk5NTNfmpk/rucvB567Gbu8v2X8D/X+AJ4PfL6+NPc48CgQVL1Tw6071NeAa4DLIuKPEfGvA71TdS/fr+pLi48DR1EFrQHLM7OvHh8IeA+1LF8JTByujvoy7ZKW42j1fOA7Lce0EOgDdl9fvZJGP0OYpK0qM+8ELqEKYxvyY+D1EbHzJu5mz5bxvYA/1uP3A39VB76BYUJm/rK1xJE2mpk9mXlWZs4CXk51+fGkiBgHfBv4DLB7Zk4GfkgV8DbX4DHU97RNbzmOVvcDRw45pvGZ+cBI9W5BTZK2IUOYpC0SES+MiA9FxPR6ek+qy4O/2ojVv0YVMr5db6cREVMi4u8j4qj1rPc39V9X7gl8APhmPf984O8i4oC6ll0j4q2bcCx/HhEviogm8ATV5b5+oAsYBywDeiPiSOB1G7vdERwcEW+uLyueQfXXpMOds/OBTw1cUo2I3ep739ZXr6TtgCFM0pZ6kupG9xsj4mmqIHEb8KENrVjf43Q4cCdwLVWQuInqMt+N61n1e8B8YAFwJdXXXpCZ3wHOobo890Rdx5GbcCx7AN+q61gI/BT4WmY+CbwfuJzqxvj/CVyxCdsd6RiOq7f3DuDN9f1hQ32+3tePIuJJqvN76Prq3cK6JG0jkTliz7wkjToRkcDMzFxcupbNFRFnUt1M//bStUgqx54wSZKkAgxhkiRJBXg5UpIkqQB7wiRJkgrY7h5gO3Xq1JwxY0bpMiRJkjZo/vz5j2TmbsMt2+5C2IwZM5g3b17pMiRJkjYoIv4w0jIvR0qSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAW0LYRFxcUQ8HBG3jbA8IuLfI2JxRNwaEXPaVYskSdJo086esEuAI9az/EhgZj2cCvxHG2uRJEkaVTrateHM/FlEzFhPk2OBr2ZmAr+KiMkR8dzMXNqumjbG/fev4Oc/v69kCWNOoxF0djbo7Gyu8xoBmVD9GFTjA+t0dDQG2w2MN5sNMnNwndZ1B9YZOjQaQSb09+ew6470WrUfftlArQPzBvY/fnzH4DBuXFV3RGzzc74+rcfRaMQ2qW/o+RwYb31PRhrf2NeI6j0YOkTEJm9rpFcYeT8jDa37H+m4hv6sjbSfjT2WTfnZXd+xbmh86PaGG1/fOWttszH7Gmn50O2sNd3fD1TnPbKfzH6gnwCyP4F+yNY2kNkH5MDBkFm1H2xD/wjzsxrvXzN/oKbBZYPr1LXWy6q6BhcMjAxuI1gzrxEQjaARVEMjiFizr/rAWzaXDK5e1772flrWa6mbTHJg/wPnlZZ2I6wfjLTdoeusqXG964ywPiPUMnL7de1+wCHsc/Ch623TTm0LYRthGnB/y/SSet46ISwiTqXqLWOvvfZqa1Hz5y/lxBP/s6370I6j0Qi6upo0m0GzWQXC1vGBX5jDhZKBX1TN5tq/uFrbDh2GCzhD2w/V1VWF4uq1SVdXFR5baxtpGFrHSAFro89X9DOuo5fOZj+djb7qtdm31nRHo2Ves4/OerqjMbCseu2o1+lo9tOMpKPRT7NuU42vPW9wOvoH12nW85ux5vWJ1eO459Fnce+jk7n3scnc8+hkHnpyIrBumI3op2tI/UOPZ7jpRiTdfU1W93awurc5ON7d16QZ/euch46Wc9B6PgbPRX0+OhprD63no/U8VeNrzkvrORo4F42B8xMt4wPzY8161fSaNo3B8RwcX2dY37J6iHWmqV5ZM91sbMIPn3ZIN/zkvTtsCNtomXkBcAHA3Llz2/qpOvzwfVi06PR27mKH09fXT09PPz09fWu/dvfSyG6a/StpsoqOXE2jfyUduYroW0X2PEP2rCJ7nyF6V0LvKqJ/FU166KCHJt00B16zm36adOcEVud4VudOrO4fz6r+8fT2N+mMbjqim05W0xHddFANTXpoRs+a8Xp7EPTFBHpiAn2NCfTGBHpjJ/qjkya9NFlNM9euIfq7oW810ddN9K8m+rtpZDdkH905ju7+caweGPqq197sopcueumkjy56cxy90UVvdtHT38Hqvk56+jvo7uuku6+Dnv4mnY0exjV7GNfsZlyjm65mD13RQ2ezh65GNx3RQ1ejh47orY+7h85GD51RHWvnwPFGL33ZbNlPJ6v7Ouju7aCnv0FH9FXrNXrpalTb6Ige+umgjy56sqq5j0766CSCatvRQzN66Yiqho6o3696fM05HzjfvTTq5Q36N/jztLX106SfJkmDpElGk/6BcRpkrHmFBl39f2R83rLWNnoZxyp2pUEPDXpp0lMdE33b/Hi2RH/LOaiOv0lG67x1z8fAckac3yBp1NuuxzMGx5NG3SYgGvW6ATTojwZ99fJqWdTbj6p7LQbWj3reutPA4PyIWHv54DotPcLRrKepX9denhEwUAOss6215g0uG9qudf6adQb3Ue+3XlCdm1pWRdA/0OuZ0D/QuVWvv+ZY1qw3cBxrZg/Zf+t6a60bLZvZ2HVa2g5tM3T9xvDrMHSd1lYj7Kd1uyO3X9uBe+494rJtoWQIewDYs2V6ej2vqIkTu9hvvymlyxhdMqHnaVi9ArqfhJ6nqqF74PXJIdNPrWnT83Q9PAO9z6x57V1ZjbOZmToa0BwHHeOr1+Y46O+DnrqWbPnlN9j33/La6IBmvW5HvX6ja814JvQ+Xh/L09D7NPSuWrPNRsea/bYOA+s3d4Lms6rxaLYc+9PQ80i9zWeqbfat3rxzsCFr1TV+mOku6O+ta3gS+lZB7+rqtb+nOh+DxzSw/sRqnb5n1rTtW10fQ6PaZuu5aHRV56L1HA89582WYWC60VkNzc41443O6rwPN7/Zsqz1NTrq6ZYhmi2vTapf+5uo52lYcS88cS+suJeOJ+5l4qpH1z6GwfHOltehNY8wHQ3o667Pbfeac9zX3XIswxxzs7M65mbn2uegOeR8DJyDRrPeX30uRvglJqk9SoawK4DTI+Iy4FBgRen7wcaswRD1OKx6DFY/Bqser14H5z0+ZHxFHbpWwOon1g4169McB50ToWti9dq5czU+blfo2Ak6d6peO3aCzglD5k2ox+v5HePr8Xpojq/nja9+gazvePtWrwmD/T1r1u+cUP/y34wf/f6+6pdgs6v65bW1ZH+13d5VdRBatSac9a1eNxw1x619XgaH8S1hqWu9/5PUFurcGaYeUA2StJnaFsIi4lLgMGBqRCwBPgF0AmTm+cAPgaOAxcAzwCntqmXM6e+DVY/CMw/BMw/Xw0Ow8hF4Zln1uvIRWFmPr3qs+uW9Pl2TYNyzYPxkGDcZdnl+FZy6dq1eB4bOSVXbwZA1ELh2rsabndvmHKxPxJqwxtStt91GExoTtt72BkSjpV5J0o6inX8decIGlidwWrv2v93JrHqenloCTz0ATz8ITz9UB62HWsYfrsJVDnPvTDRgwtQ1w7NfWL2Of3ZLwHoWjH9WFbTGTa7Hd928niFJkrTZ/M27rfSuhifvgyf+sGZ48j54ckk1PLWkumQ4VMcE2Gl32Hl32HVveO6h1fTAvJ2eU08/pwpU3tMhSdJ2wRC2qbL+3pbsX3u8bzU8ef/aIWvFvfBkPf70g2tvJxqw8/Ng0p6w24thn6Ng4nSYNB0mToOd96jCVefO3tsjSdIYZAgbKrO67Pf43bDi9/B4y7Di99XlwI3R7IJJe1X3Vu19FOwyoxofGCZOGx33T0mSpCIMYUMtuhyuPL5lRlS9U5NfAPv8d5j4vPr+qTXfUwNR/Zn3pOlrQtbOe3hpUJIkjcgQNtRzD4U///cqdE1+QdWD1TGudFWSJGmMMYQNtesMmPO+0lVIkqQxzutlkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgpoawiLiCMiYlFELI6Ijw6zfK+IuD4ifh0Rt0bEUe2sR5IkabRoWwiLiCZwHnAkMAs4ISJmDWn2MeDyzDwIOB74X+2qR5IkaTRpZ0/YIcDizLw7M7uBy4Bjh7RJYJd6fFfgj22sR5IkadRoZwibBtzfMr2kntfqTODtEbEE+CHwvuE2FBGnRsS8iJi3bNmydtQqSZK0TZW+Mf8E4JLMnA4cBXwtItapKTMvyMy5mTl3t9122+ZFSpIkbW3tDGEPAHu2TE+v57V6F3A5QGb+P2A8MLWNNUmSJI0K7QxhNwMzI2LviOiiuvH+iiFt7gNeAxAR+1OFMK83SpKkMa9tISwze4HTgWuAhVR/BXl7RJwdEcfUzT4E/GVE/Aa4FPiLzMx21SRJkjRadLRz45n5Q6ob7lvnfbxl/A7gFe2sQZIkaTQqfWO+JEnSDskQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKaGsIi4gjImJRRCyOiI+O0OZtEXFHRNweEd9oZz2SJEmjRUe7NhwRTeA84LXAEuDmiLgiM+9oaTMT+DvgFZn5WEQ8p131SJIkjSbt7Ak7BFicmXdnZjdwGXDskDZ/CZyXmY8BZObDbaxHkiRp1GhnCJsG3N8yvaSe12o/YL+I+EVE/CoijhhuQxFxakTMi4h5y5Yta1O5kiRJ207pG/M7gJnAYcAJwIURMXloo8y8IDPnZubc3XbbbRuXKEmStPW1M4Q9AOzZMj29ntdqCXBFZvZk5j3A76hCmSRJ0pjWzhB2MzAzIvaOiC7geOCKIW2+S9ULRkRMpbo8eXcba5IkSRoV2hbCMrMXOB24BlgIXJ6Zt0fE2RFxTN3sGmB5RNwBXA/8TWYub1dNkiRJo0VkZukaNsncuXNz3rx5pcuQJEnaoIiYn5lzh1tW+sZ8SZKkHZIhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVsN4QFhGvj4i3DDP/LRHx2vaVJUmSNLZtqCfs48BPh5l/A3D2Vq9GkiRpB7GhEDYuM5cNnZmZjwA7t6ckSZKksW9DIWyXiOgYOjMiOoEJ7SlJkiRp7NtQCPtP4MKIGOz1ioiJwPn1MkmSJG2GDYWwjwEPAX+IiPkRMR+4B1hWL5MkSdJmWOdSY6vM7AU+GhFnAfvWsxdn5sq2VyZJkjSGbegrKmZGxPeAm4G/Bx41gEmSJG25DV2OvBj4AfA/gFuAL7S9IkmSpB3Aei9HApMy88J6/NMRcUu7C5IkSdoRbCiEjY+Ig4CopydExJyBhZlpKJMkSdoMGwphS4HPsiaEPQh8pmX5q9tRlCRJ0li3oRD2t8D9mbkUICJOpro/7F7gzLZWJkmSNIZt6Mb884HVABHxZ8A/A18BVgAXtLc0SZKksWtDPWHNzHy0Hj8OuCAzvw18OyIWtLc0SZKksWtDPWHNlmdHvgb4ScuyDQU4SZIkjWBDQepS4KcR8QiwEvg5QETsS3VJUpIkSZthQ48t+lREXAc8F/hRZma9qAG8r93FSZIkjVUbvKSYmb8aZt7v2lOOJEnSjmFD94RJkiSpDQxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVEBbQ1hEHBERiyJicUR8dD3t/kdEZETMbWc9kiRJo0XbQlhENIHzgCOBWcAJETFrmHaTgA8AN7arFkmSpNGmnT1hhwCLM/PuzOwGLgOOHabdJ4FzgFVtrEWSJGlUaWcImwbc3zK9pJ43KCLmAHtm5pXr21BEnBoR8yJi3rJly7Z+pZIkSdtYsRvzI6IB/BvwoQ21zcwLMnNuZs7dbbfd2l+cJElSm7UzhD0A7NkyPb2eN2AScCBwQ0TcC7wUuMKb8yVJ0o6gnSHsZmBmROwdEV3A8cAVAwszc0VmTs3MGZk5A/gVcExmzmtjTZIkSaNC20JYZvYCpwPXAAuByzPz9og4OyKOadd+JUmStgcd7dx4Zv4Q+OGQeR8foe1h7axFkiRpNPEb8yVJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAtoawiLiiIhYFBGLI+Kjwyz/64i4IyJujYjrIuL57axHkiRptGhbCIuIJnAecCQwCzghImYNafZrYG5mvhj4FvCv7apHkiRpNGlnT9ghwOLMvDszu4HLgGNbG2Tm9Zn5TD35K2B6G+uRJEkaNdoZwqYB97dML6nnjeRdwFXDLYiIUyNiXkTMW7Zs2VYsUZIkqYxRcWN+RLwdmAt8erjlmXlBZs7NzLm77bbbti1OkiSpDTrauO0HgD1bpqfX89YSEYcD/wC8KjNXt7EeSZKkUaOdPWE3AzMjYu+I6AKOB65obRARBwH/GzgmMx9uYy2SJEmjSttCWGb2AqcD1wALgcsz8/aIODsijqmbfRqYCPzfiFgQEVeMsDlJkqQxpZ2XI8nMHwI/HDLv4y3jh7dz/5IkSaPVqLgxX5IkaUdjCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBHaUL2Bp6enpYsmQJq1atKl3Kdm/8+PFMnz6dzs7O0qVIkjSmjYkQtmTJEiZNmsSMGTOIiNLlbLcyk+XLl7NkyRL23nvv0uVIkjSmjYnLkatWrWLKlCkGsC0UEUyZMsUeRUmStoExEcIAA9hW4nmUJGnbGDMhTJIkaXtiCNsKli9fzuzZs5k9ezZ77LEH06ZNG5zu7u7eqG2ccsopLFq0aKP3uXTpUo466ihe8pKXMGvWLI455hgAFi9ezIQJE5g9ezazZs3itNNOIzMBuPPOOznyyCOZOXMmc+bM4fjjj+fhhx/e9AOWJElbbEzcmF/alClTWLBgAQBnnnkmEydO5MMf/vBabTKTzKTRGD73fvnLX96kfX7sYx/j6KOP5rTTTgPg1ltvHVz2J3/yJyxYsICenh4OO+wwvv/973P44Ydz9NFH84UvfIGjjjoKgOuuu47ly5fznOc8Z5P2LUmSttyYC2FnnHE1CxY8uFW3OXv2Hpx77hGbvN7ixYs55phjOOigg/j1r3/Ntddey1lnncUtt9zCypUrOe644/j4xz8OwCtf+Uq++MUvcuCBBzJ16lTe8573cNVVV7HTTjvxve99b52gtHTpUqZPnz44/eIXv3id/Xd2dvKyl72MxYsX8+CDD/KqV71qMIABvOY1r9nkY5IkSVuHlyPb7M477+SDH/wgd9xxB9OmTeNf/uVfmDdvHr/5zW+49tprueOOO9ZZZ8WKFbzqVa/iN7/5DS972cu4+OKL12lz+umnc/LJJ/PqV7+af/qnf2Lp0qXrtHn66af5yU9+wote9CJuu+02Dj744LYcoyRJ2nRjridsc3qs2ukFL3gBc+fOHZy+9NJL+dKXvkRvby9//OMfueOOO5g1a9Za60yYMIEjjzwSgIMPPpif//zn62z3qKOO4ve//z1XX301V111FQcddBC33347AIsWLWL27Nk0Gg3e9KY38drXvpbvf//7bTxKSZK0qcZcCBttdt5558Hxu+66i89//vPcdNNNTJ48mbe//e3DfidXV1fX4Hiz2aS3t3fYbU+ZMoUTTzyRE088kSOOOIL/+q//4oADDhi8J6zVAQccwI033riVjkqSJG0pL0duQ0888QSTJk1il112YenSpVxzzTWbva3rrruOlStXDm73nnvuYa+99hqx/Tve8Q5uuOEGrr766sF5119/PQsXLtzsGiRJ0uYzhG1Dc+bMYdasWbzwhS/kpJNO4hWveMVmb+vmm29mzpw5vPjFL+blL385733veznooINGbL/TTjvxgx/8gM997nPMnDmTWbNmceGFFzJ16tTNrkGSJG2+GPgOqe3F3Llzc968eWvNW7hwIfvvv3+hisYez6ckSVtHRMzPzLnDLbMnTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCNsKli9fzuzZs5k9ezZ77LEH06ZNG5zu7u7e6O1cfPHFPPjg8A8f/8UvfsGhhx7K7Nmz2X///fnkJz8JwEUXXcRuu+02OL/1OZNXXnklBx98MAcccACzZ8/mb//2b7fsQCVJ0lbjY4u2gilTpgw+JujMM45zl9kAAAytSURBVM9k4sSJfPjDH97k7Vx88cXMmTOHPfbYY51lJ598Mt/97nc58MAD6evrY9GiRYPLTjzxRM4991wefPBBDjzwQI455hgeeOABzjjjDK688kr2228/+vr6uOCCCzb/ICVJ0lY19kLY9WfAwws23G5TPGc2/Pm5m7XqV77yFc477zy6u7t5+ctfzhe/+EX6+/s55ZRTWLBgAZnJqaeeyu67786CBQs47rjjmDBhAjfddNNaz5BctmzZYDhrNpvrPPQbYI899mDGjBncd999fOYzn+Ef//Ef2W+//QbXee9737tZxyBJkrY+L0e20W233cZ3vvMdfvnLX7JgwQJ6e3u57LLLmD9/Po888gi//e1vue222zjppJM47rjjmD17Nt/85jdZsGDBWgEM4IwzzmDmzJm8+c1v5sILL2T16tXr7G/x4sX84Q9/YJ999uG2227j4IMP3laHKkmSNtHY6wnbzB6rdvjxj3/MzTffzNy51dMKVq5cyZ577snrX/96Fi1axPvf/36OPvpoXve6121wW2eddRbveMc7+NGPfsRXv/pVvvnNb/LjH/8YgK9//evccMMNjBs3josuuojJkye39bgkSdKWG3shbBTJTN75zncO3kTf6tZbb+Wqq67ivPPO49vf/vZG3a+17777su+++/Lud7+bqVOnsmLFCmDNPWGtDjjgAObPn88BBxywdQ5GkiRtVV6ObKPDDz+cyy+/nEceeQSo/oryvvvuY9myZWQmb33rWzn77LO55ZZbAJg0aRJPPvnksNu68sorGXjY+l133cW4ceOYNGnSiPv+yEc+wic/+UkWL14MQF9fH+eff/7WPDxJkrQF7Alroxe96EV84hOf4PDDD6e/v5/Ozk7OP/98ms0m73rXu8hMIoJzzjkHgFNOOYV3v/vdw96Yf8kll/DXf/3XTJgwgc7OTr7xjW/QaIycoQ866CA++9nP8ra3vY1Vq1YBcOyxx7b3gCVJ0kaLgd6V7cXcuXNz3rx5a81buHAh+++/f6GKxh7PpyRJW0dEzM/MucMt83KkJElSAYYwSZKkAsZMCNveLquOVp5HSZK2jTERwsaPH8/y5csNEFsoM1m+fDnjx48vXYokSWPemPjryOnTp7NkyRKWLVtWupTt3vjx45k+fXrpMiRJGvPGRAjr7Oxk7733Ll2GJEnSRmvr5ciIOCIiFkXE4oj46DDLx0XEN+vlN0bEjHbWI0mSNFq0LYRFRBM4DzgSmAWcEBGzhjR7F/BYZu4LfA44p131SJIkjSbt7Ak7BFicmXdnZjdwGTD0K9uPBb5Sj38LeE1ERBtrkiRJGhXaeU/YNOD+luklwKEjtcnM3ohYAUwBHmltFBGnAqfWk09FxKKtXOvUofvUmOb7vWPx/d5x+F7vWLaX9/v5Iy3YLm7Mz8wLgAvatf2ImDfSIwU09vh+71h8v3ccvtc7lrHwfrfzcuQDwJ4t09PrecO2iYgOYFdgeRtrkiRJGhXaGcJuBmZGxN4R0QUcD1wxpM0VwMn1+FuAn6TfuCpJknYAbbscWd/jdTpwDdAELs7M2yPibGBeZl4BfAn4WkQsBh6lCmoltO1Sp0Yl3+8di+/3jsP3esey3b/fYceTJEnStjcmnh0pSZK0vTGESZIkFTDmQ1hE7BkR10fEHRFxe0R8oJ7/7Ii4NiLuql+fVc+PiPj3+lFKt0bEnLJHoM0REc2I+HVE/KCe3rt+NNbi+lFZXfV8H521nYuIyRHxrYi4MyIWRsTL/HyPXRHxwfrf8tsi4tKIGO/ne+yIiIsj4uGIuK1l3iZ/niPi5Lr9XRFx8nD7Gg3GfAgDeoEPZeYs4KXAafXjkz4KXJeZM4Hr6mmoHrM0sx5OBf5j25esreADwMKW6XOAz9WPyHqM6pFZ4KOzxoLPA1dn5guBl1C9736+x6CImAa8H5ibmQdS/dHX8fj5HksuAY4YMm+TPs8R8WzgE1RfEH8I8ImB4DbajPkQlplLM/OWevxJqn+gp7H2I5O+AryxHj8W+GpWfgVMjojnbuOytQUiYjpwNHBRPR3Aq6kejQXrvt8+Oms7FRG7An9G9ZfWZGZ3Zj6On++xrAOYUH+35E7AUvx8jxmZ+TOqb0totamf59cD12bmo5n5GHAt6wa7UWHMh7BWdVf0QcCNwO6ZubRe9CCwez0+3OOWpm2jErV1nAt8BOivp6cAj2dmbz3d+p6u9egsYODRWdo+7A0sA75cX36+KCJ2xs/3mJSZDwCfAe6jCl8rgPn4+R7rNvXzvN18zneYEBYRE4FvA2dk5hOty+oviPW7OsaAiHgD8HBmzi9di7aJDmAO8B+ZeRDwNGsuVQB+vseS+pLSsVTh+3nAzozSHg61x1j7PO8QISwiOqkC2Ncz8z/r2Q8NXIaoXx+u52/M45Y0er0COCYi7gUuo7pM8XmqbuqBLydufU99dNb2bQmwJDNvrKe/RRXK/HyPTYcD92TmsszsAf6T6jPv53ts29TP83bzOR/zIay+/v8lYGFm/lvLotZHJp0MfK9l/kn1X128FFjR0g2qUS4z/y4zp2fmDKobdn+SmScC11M9GgvWfb99dNZ2KjMfBO6PiD+pZ70GuAM/32PVfcBLI2Kn+t/2gffbz/fYtqmf52uA10XEs+re09fV80adMf+N+RHxSuDnwG9Zc4/Q31PdF3Y5sBfwB+Btmflo/cH+IlUX9zPAKZk5b5sXri0WEYcBH87MN0TEPlQ9Y88Gfg28PTNXR8R44GtU9wo+ChyfmXeXqlmbLiJmU/0RRhdwN3AK1X8w/XyPQRFxFnAc1V++/xp4N9X9Pn6+x4CIuBQ4DJgKPET1V47fZRM/zxHxTqrf9QCfyswvb8vj2FhjPoRJkiSNRmP+cqQkSdJoZAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSVtFRGREfLZl+sMRceZW2vYlEfGWDbfc4v28NSIWRsT17d7XBuq4NyKmlqxBUvsZwiRtLauBN4+28NDyTeob413AX2bmn7erHkkaYAiTtLX0AhcAHxy6YGhPVkQ8Vb8eFhE/jYjvRcTdEfEvEXFiRNwUEb+NiBe0bObwiJgXEb+rnxFKRDQj4tMRcXNE3BoRf9Wy3Z9HxBVU36g+tJ4T6u3fFhHn1PM+DrwS+FJEfHpI++dGxM8iYkG9zn+r5/9HXdPt9ZeIDrS/NyL+uW4/LyLmRMQ1EfH7iHhPS40/i4grI2JRRJwfEev8mxwRb6/Px4KI+N/1MTfrc3pbfRzrnHNJo9+m/A9RkjbkPODWiPjXTVjnJcD+VN9ofjdwUWYeEhEfAN4HnFG3mwEcArwAuD4i9gVOonpUyZ9GxDjgFxHxo7r9HODAzLyndWcR8TzgHOBg4DHgRxHxxsw8OyJeTfWUhaHfov8/gWsy81MR0QR2quf/Q/3N3U3guoh4cWbeWi+7LzNnR8TngEuonnE4HrgNOL9ucwgwi+pbwK8G3kz1/MuBWven+nb4V2RmT0T8L+BE4HZgWmYeWLebvDEnWtLoYk+YpK0mM58Avgq8fxNWuzkzl2bmauD3wECI+i1V8BpweWb2Z+ZdVGHthVTPhDspIhZQPYpsCjCzbn/T0ABW+1Pghvoh0L3A14E/21CNwCn1PW4vyswn6/lvi4hbqB6VcwBVoBpwRctx3JiZT2bmMmB1S2i6KTPvzsw+4FKqnrhWr6EKizfXx/gaYJ/6+PeJiC9ExBHAExuoX9IoZE+YpK3tXOAWoPVZbb3U/+mrL7l1tSxb3TLe3zLdz9r/Rg19xloCAbwvM9d6OG/93NCnN6/8dWXmzyLiz4CjgUsi4t+onkn7YeBPM/OxiLiEqqdrQOtxDD3GgeMa7phaBfCVzPy7oTVFxEuA1wPvAd4GvHNTj0tSWfaESdqqMvNRqoftvqtl9r1UPToAxwCdm7Hpt0ZEo75PbB9gEXAN8N6I6ASIiP0iYucNbOcm4FURMbW+jHgC8NP1rRARzwceyswLqR4WPgfYhSrorYiI3YEjN+OYDomIvetgehzwX0OWXwe8JSKeU9fx7Ih4fv3HD43M/DbwsboeSdsZe8IktcNngdNbpi8EvhcRv6G692lzeqnuowpQuwDvycxVEXER1SXLWyIigGXAG9e3kcxcGhEfBa6n6mm6MjO/t4F9Hwb8TUT0AE8BJ2XmPRHxa+BO4H7gF5txTDcDXwT2rev5zpBa74iIj1Hdt9YAeoDTgJXAl1tu5F+np0zS6BeZQ3u/JUntVl8y/XBmvqF0LZLK8HKkJElSAfaESZIkFWBPmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBXw/wEi+mT95BUpqAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "trn_score_spc = exponential_moving_average(score_trn_list[:,1], 0.1)\n",
    "tst_score_spc = exponential_moving_average(score_tst_list[:,1], 0.1)\n",
    "\n",
    "\n",
    "plt.title(\"SPC per samples\")\n",
    "plt.xlabel('Number of samples')\n",
    "plt.ylabel('SPC')\n",
    "plt.plot(train_size, trn_score_spc, label='Train SPC',color='navy')\n",
    "plt.plot(train_size, tst_score_spc, label='Test SPC',color=\"darkorange\")\n",
    "plt.legend(loc=\"best\")\n",
    "plt.ylim(0,1.05)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Precision PPV = TP / (TP + FP):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "trn_score_ppv = exponential_moving_average(score_trn_list[:,2], 0.1)\n",
    "tst_score_ppv = exponential_moving_average(score_tst_list[:,2], 0.1)\n",
    "\n",
    "plt.title(\"PPV per samples\")\n",
    "plt.xlabel('Number of samples')\n",
    "plt.ylabel('PPV')\n",
    "plt.plot(train_size, trn_score_ppv, label='Train PPV',color='navy')\n",
    "plt.plot(train_size, tst_score_ppv, label='Test PPV',color=\"darkorange\")\n",
    "plt.legend(loc=\"best\")\n",
    "plt.ylim(0,1.05)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Negative Predictive Value NPV = TN / (TN + FN):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "trn_score_npv = exponential_moving_average(score_trn_list[:,3], 0.1)\n",
    "tst_score_npv = exponential_moving_average(score_tst_list[:,3], 0.1)\n",
    "\n",
    "plt.title(\"NPV per samples\")\n",
    "plt.xlabel('Number of samples')\n",
    "plt.ylabel('NPV')\n",
    "plt.plot(train_size, trn_score_npv, label='Train NPV',color='navy')\n",
    "plt.plot(train_size, tst_score_npv, label='Test NPV',color=\"darkorange\")\n",
    "plt.legend(loc=\"best\")\n",
    "plt.ylim(0,1.05)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### False Positive Rate FPR = FP / (FP + TN)):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "trn_score_fpr = exponential_moving_average(score_trn_list[:,4], 0.1)\n",
    "tst_score_fpr = exponential_moving_average(score_tst_list[:,4], 0.1)\n",
    "\n",
    "plt.title(\"FPR per samples\")\n",
    "plt.xlabel('Number of samples')\n",
    "plt.ylabel('FPR')\n",
    "plt.plot(train_size, trn_score_fpr, label='Train FPR',color='navy')\n",
    "plt.plot(train_size, tst_score_fpr, label='Test FPR',color=\"darkorange\")\n",
    "plt.legend(loc=\"best\")\n",
    "plt.ylim(0,1.05)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### False Discovery rate FDR = FP / (FP + TP):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "trn_score_fdr = exponential_moving_average(score_trn_list[:,5], 0.1)\n",
    "tst_score_fdr = exponential_moving_average(score_tst_list[:,5], 0.1)\n",
    "\n",
    "plt.title(\"FDR per samples\")\n",
    "plt.xlabel('Number of samples')\n",
    "plt.ylabel('FDR')\n",
    "plt.plot(train_size, trn_score_fdr, label='Train FDR',color='navy')\n",
    "plt.plot(train_size, tst_score_fdr, label='Test FDR',color=\"darkorange\")\n",
    "plt.legend(loc=\"best\")\n",
    "plt.ylim(0,1.05)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### False Negative rate FNR = FN / (FN + TP):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "trn_score_fnr = exponential_moving_average(score_trn_list[:,6], 0.1)\n",
    "tst_score_fnr = exponential_moving_average(score_tst_list[:,6], 0.1)\n",
    "\n",
    "plt.title(\"FNR per samples\")\n",
    "plt.xlabel('Number of samples')\n",
    "plt.ylabel('FNR')\n",
    "plt.plot(train_size, trn_score_fnr, label='Train FNR',color='navy')\n",
    "plt.plot(train_size, tst_score_fnr, label='Test FNR',color=\"darkorange\")\n",
    "plt.legend(loc=\"best\")\n",
    "plt.ylim(0,1.05)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Accuraccy ACC = (TP + TN) / (P + N): "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmEAAAG5CAYAAADGcOOUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nO3de5xcdZ3n/9enq7vTCQlBQiCaAEFAIYLGEKMo4xUR1B+wrqiM/nBAZdwZrzPo4oyriDOul1W8wOriCDiOE2RhFB4jDt5QmVGEBCNKEAmQCVEuIVwSLklf6rN/nNNJdafTIZfKt9N5PR+PepxrnfM553RVvet7TteJzESSJEk7V0fpAiRJknZHhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSWNQRCyPiGNL1yGpfQxhknaIiPhJRDwUERNGmLYgIq6OiIcj4sGIuCEiTm+ZvmdEfD4iVkTEoxFxRz28z87dCknaeQxhkrZbRMwG/gRI4MRh044Gfgz8FDgEmAb8N+CEeno38CPgWcDxwJ7A0cBqYEGb6+5s5/IlaTSGMEk7wmnA9cAlwFuHTfsM8PXM/FRmPpCVxZn5hpbnHgD8l8xcmpnNzLw/Mz+emVePtLKIyIh4T0TcGREPRMRnIqKjZfoZEXFr3TJ3TUQcOOy5fxkRtwO3j7Dsnoj4p4hYXbfc3RgR+9XTTq+Xu7Ze95+3PO+lEbEyIj4YEfdHxD0RcXJEvDoifl+3AP5Ny/znRMTlEfGtenk3RcRzNrO9HRFxdt1CuDoiLouIvbdUr6SxzRAmaUc4Dfhm/XhVS2iZRNWqdfkozz0W+LfMfHQr1/lfgPnAPOAk4Ix6nScBfwO8DpgOXAcsHPbck4HnA3NGWO5bganA/lStdu8Enqin3Q+8lqq17nTgvIiY1/LcGUAPMBP4CPBV4C3AUVQthf8jIg5qmf8k4P8CewP/DHwnIrpGqOnddc0vAZ4GPARc8CTqlTSGGcIkbZeIOAY4ELgsMxcDdwB/Wk9+CtX7zD2jLGLaFqZvzqcy88HMXAF8Hji1Hv9O4H9m5q2Z2Q98Apjb2hpWT38wM0cKK311TYdk5kDdarcGIDO/m5l31K15PwW+TxWuWp/795nZB1wK7AN8ITPXZuYtwFKgtbVrcWZeXs//OaoA94IRanon8LeZuTIz1wPnAK+vT6dutl5JY5shTNL2eivw/cx8oB7+ZzaeknwIaAJPHeX5q7cwfXPubun/T6oWIqgC4RfqU3MPAw8CQdU6NdJzh/sGcA1waUT8MSI+Pdg6FREnRMT19anFh4FXUwWtDduSmQN1/2DAu69l+hPA5JHqyMwmsLJlO1odCHy7ZZtuBQaA/UarV9LYZgiTtM0iYiLwBuAlEXFvRNwLvB94TkQ8JzMfB34B/NdRFvNDqlOYe2zl6vdv6T8A+GPdfzfw55m5V8tjYmb+vGX+3NxCM7MvMz+WmXOAF1Kdfjyt/q/PK4D/BeyXmXsBV1MFvG21YRvqa9pmtWxHq7uBE4ZtU09m/mFz9W5HTZJ2EkOYpO1xMlWLzBxgbv04nOo6rMEg8EHgzyLiAxExDSAinhMRl9bTv0EVMq6IiMPqi9CnRcTfRMSrR1n3ByLiKRGxP/Be4Fv1+K8AH4qIZ9XrmhoRpzzZDYqIl0XEkRHRANZQne5rAt3ABGAV0B8RJwDHPdnlbsZREfG6+rTi+4D1VP/gMNxXgL8fPKUaEdPra99Gq1fSGGcIk7Q93gpcnJkrMvPewQdwPvDmiOisW6BeXj/ujIgHgQupWpGor3E6Fvgd8AOqIHED1Wm+X46y7iuBxcAS4LvA1+rlfRv4FNXpuTXAb6l/DuNJmkH1jwRrqE77/RT4RmauBd4DXEZ1mvVPgau2Yrmb24Y31sv7/4HX1deHDfeFel3fj4i1VEHt+aPVu511SdoJInOzrfKSNCZFRAKHZuay0rVsq4g4h+pi+reUrkVSGbaESZIkFWAIkyRJKsDTkZIkSQXYEiZJklTALnfz2n322Sdnz55dugxJkqQtWrx48QOZOX2kabtcCJs9ezaLFi0qXYYkSdIWRcR/bm6apyMlSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqYC2hbCIuCgi7o+I325mekTEFyNiWUTcHBHz2lWLJEnSWNPOlrBLgONHmX4CcGj9OBP4chtrkSRJGlM627XgzPxZRMweZZaTgH/MzASuj4i9IuKpmXlPu2p6Mu6++xGuu25FyRJ2WY1G0N3doLu7QVdXo6W/g87O6tHV1ai71TBAJmTmkG5HR2x4Tuv8HR1Bs5kjPicCOjs7aDQ2zrs5mblhOVX3yfdvrtvREfT0dDJhQoMJEzpHXb+2rHX/th7zzfVDdQyGPyLiSR/Xzf1ttXY3t57BdW3sZ8i4CIjY2B1tW2HovMOfM7yujc/ZOP/guktp3Z/DaxtpW7b3GI20jsHu1h7j0Y59tZ6h+3hwPSMd58HuSO8hG5c3dN5Go3ov6enppNHwqqHxrG0h7EmYCdzdMryyHrdJCIuIM6layzjggAPaWtTixffw5jf/S1vXoZ0jAhqNjcGt9U1wZ+jq6mDChE66uxs0GtUb9GA9g4/BD4WN9W380NncB33r/CM9BpcxuJzRxg0G566uKiwPdjs7O4bUsqV1Nps55INpeCgZbRmbq21n6ogmXY0BugcfnQN0djTpajTp6hiouo2qu2bdBFY8PJXHe7u3a52tH85b85yt3TcRbPhS1NXVseHLUXd39YXoyRzbzR2zkb7MDD1+We/Har92NZp1d+Pw4PTB/T3ScOu4zo4mnS3jBx8R0DvQoG+gg77BbrPqPtbbzZp1E6rH+gk8Uvf3DTQ2HNfhx3m04da/jc5Gk66WmhqxaX1AVVOzpbaW4f5mR/UY6KCv2UEzg0YkXY0Berph4oRgUk8wqQd6upOuRvXobLTso86ga9IUeibvycSpT2HyXnsxZdo09po+jclTJ8NAH83+XrK/l+ZAHwz0ka2PZv+GcTT7q+FmP2Q/NAeq/mY/5ABkP1H3x+BwDrQMDxDZT2QTGByXBANENgmq8WTSmxNYNzCR9c0enhiYWD36J9BM6GCAoEkH/Rv6g6Sv2c267KG32UNvTmR9cwK99FTvafRX82c/wQANqm5H/di4vCYdMcBBx7yGF5/y+q17Ue1AJUPYk5aZFwIXAsyfP7+tb8/HHvt0brvtXe1cxbiUmQwMJL29A/T1DdDbu/HR39+kr69Jf3+z7t84Dkb+9prJJvP39zcZGMjNftvMhIGBjfMNPqfZTBqNoS0SI7VQDF/uaK0YI3WbzWT9+gHWretn/fr+ulvtg2YzGRhotnyQVbUODSwbA0y1TxnywTcwkBu2ZdMWmCpwblwGdEY/Exp9dHX0QXSSHZ006SI7uiAaRAQDA9VxGDxufX0bj1VrTV0d/fR09tLd0Uuj0SCjk4wusqNRdaNBBwM0ch2d9NKZ62mwnkaup5P1TGj00t3Rx4RGH90d1XK6OqpxXdFHV0dv1Y0+OqOXzrq/Ef00oo9Oqv6q20eDPhr000lvNQ99JNCki/7sYoDq0Z9dJFHVRh+dUb1BN6ie0xF99Zt2c6v/5texJ2tjBmuZwRr247HciyYNBrJBMzsYyA6aWX2gdmRfvT9669p7adBb7asN21pvP310xADNbFQfG9k5ZLmNGKCz3gedG/ZPP0GT9UxiXe7BupzM+tyDJ5qTWN+cWH1IDqwnmr105MZHg+rvo7ujn86O/vpY9NEV1XBn9NMZg+up+pt00MwGTepH3Z8Ejaj284ZjRT8ROzlRq9IH3Fs/2rWKgSo8DjSDgTpIDmTQP1D9/Q+OH8jWbtSvjSATJnX3sd+EXiZP6GXKpF46G1v/WtweN9y/J7B7hrA/APu3DM+qxxU1eXI3z3jGtNJllDV4bu/JGOiF3rXDHmuq7vo10Le2Wl5nD3ROhEZPS/8EGFgPfY9uuoy+R6HvsY3d3rqb/RuX0drtqpfdmDB0WmdPVWfvoxvX0/doNTywDjq6odFdPW+k7obpdX8E9D1e19ZS3/BHf8s8A+s3LnP4+jq6qkej7nZ0bhw3ZLjuRke17N5Hof+xjdvV91hVV/8T1fT+JyBHezOLoeuMznodg+uJahl9j1fLa/Zv95/VqBrdw47fhGHHYuLQ4U2OW90q1b8emr3VPh/sbw5snKf1mHZ0tYzrGja+a+hxaHRV+6jRBesfgTUr6FlbPaavWQFrf12Nr3cto718Ojpbtm/Y9nb2QGPPap3Nfmj21d11G1siGl31Ppg0dB8Q9WtvDaz/z6qe3keq1yjU2zG4nglDa2hMgMaUEaZ1D93XHV1AbqxlQ0tJ3UIy/PWyYX9OqPdp97DuZv7OR+x2Dh0e8jdbf5Q1+6rtbVYtPRuG+x+v98ualn20pvr72OQ1+CSHR62xtb5GVdtgPYOPweEcqPoH92Ozr/qbHb69rcsc/Hscvv31e9PAujU88sBqHl71AI888CDrHn98Q93RqPZ9NLqIjrrb2U00OutH1d/R6KKjs6se10VHRyfR2UlHRycdXd10NBqbtHoPfkGcMML44V+CW7/kbpAJ/euq97TB10o0hnUH35taPhcG3wMH39cGX6tD3jsbIx6fBZ0Tt/vta3uUDGFXAe+KiEuB5wOPlL4ebNzIrMLGutXwxGpY91D1Zrzu4aq7/uHqDXr9wy3jWsb3rqk+7DsntjwmVd0c2BiyetdWH3Y7WqMbuibXjz2gu+5Oflr1QhpYXwWo3jXw+H3Vi7b/iY3j+9eNXFdnz8bldtcfOM2++nm9w7p1P6N8i+/sgc49qtpaH3vMGDrcmFB/MAxfR+/QN+T+dS1vwn3DPoT7Nn7QdU4aum8m7Veva9LQYzU43JhQn07obfkgGOxv+RAdsp7mxmV07VGvc/BvIIfVVdff6Nr499Lo2djfNcK4IY8J1d/brm6g5RgN6TaHBq3BD+WdWVdHY3zs4y3qKV3A5nXthHVMmApAA9j7qbD3TljlDhVRvV90bSEYDb63Ttp359TVRm0LYRGxEHgpsE9ErAQ+Sv1nmJlfAa4GXg0sAx4HTm9XLbu8Zj888QA8fj88vgqeWFV3hw1vCF2rt9ByEdWLdcJUmLBX1Z1yAEx/NnTX48mWlpWW1pVoVAGme8+6+yT6o6NexrqWwFT3Nya0zF8/Bls1tkdmHXbWVf3dkzd+W9wazYGWANVbfaB2TapCyc7+MNXY1qhbS8aasViTJKC9/x156hamJ/CX7Vr/mJdZtVA99kd4tH4M9j92Tx247q8C1rqHRl5GdEDPNJg0HSbuA3sfVg1PnFZ394GevaHnKRvD1oS9qkCys78V19/QdpqIqoWlc8L2LaejAR0TgbJN1pKk8WeXuDB/l9S7FtasgLV3b+axsmpdGq7nKbDHU6tTTNPnVgFr0r4wcXodtlqGe55ia4wkSbsoQ9homv3w6B/gkbvgkeVV94lVw2aqLyps9lXzrq2D1+BFuq3zTX4qTNm/Ou339NdU/ZNnwh5Pq6532uOpWz4XLkmSxgVD2HArr4OffxTWLK/C1JBrq6I6vbfhVF7rr+01qiA19ekw6yXVNVZT9q8ee+5fBS2vzZAkSTVD2HDRqC4Yf+rRcNipsOdBMHU27Dkb9jxgx1w0LkmSdnuGsOFmvhD+9Oelq5AkSePc7vDDMZIkSWOOIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAW0NYRFxfETcFhHLIuLsEaYfEBHXRsSvIuLmiHh1O+uRJEkaK9oWwiKiAVwAnADMAU6NiDnDZvswcFlmPhd4E/C/21WPJEnSWNLOlrAFwLLMvDMze4FLgZOGzZPAnnX/VOCPbaxHkiRpzGhnCJsJ3N0yvLIe1+oc4C0RsRK4Gnj3SAuKiDMjYlFELFq1alU7apUkSdqpSl+YfypwSWbOAl4NfCMiNqkpMy/MzPmZOX/69Ok7vUhJkqQdrZ0h7A/A/i3Ds+pxrd4GXAaQmb8AeoB92liTJEnSmNDOEHYjcGhEHBQR3VQX3l81bJ4VwCsAIuJwqhDm+UZJkjTutS2EZWY/8C7gGuBWqv+CvCUizo2IE+vZ/hp4R0T8GlgI/FlmZrtqkiRJGis627nwzLya6oL71nEfaelfCryonTVIkiSNRaUvzJckSdotGcIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgFtDWERcXxE3BYRyyLi7M3M84aIWBoRt0TEP7ezHkmSpLGis10LjogGcAHwSmAlcGNEXJWZS1vmORT4EPCizHwoIvZtVz2SJEljSTtbwhYAyzLzzszsBS4FTho2zzuACzLzIYDMvL+N9UiSJI0Z7QxhM4G7W4ZX1uNaPQN4RkT8R0RcHxHHj7SgiDgzIhZFxKJVq1a1qVxJkqSdp/SF+Z3AocBLgVOBr0bEXsNnyswLM3N+Zs6fPn36Ti5RkiRpx2tnCPsDsH/L8Kx6XKuVwFWZ2ZeZdwG/pwplkiRJ41o7Q9iNwKERcVBEdANvAq4aNs93qFrBiIh9qE5P3tnGmiRJksaEtoWwzOwH3gVcA9wKXJaZt0TEuRFxYj3bNcDqiFgKXAt8IDNXt6smSZKksSIys3QNW2X+/Pm5aNGi0mVIkiRtUUQszsz5I00rfWG+JEnSbskQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFbDaERURPREwfYfz0iOhpb1mSJEnj22gtYV8E/mSE8ccA57WnHEmSpN3DaCHsqMz8l+EjM/PbwIvbV5IkSdL4N1oIm7SNz5MkSdIWjBam7o+IBcNHRsTzgFXtK0mSJGn86xxl2geAyyLiEmBxPW4+cBrwpjbXJUmSNK5ttiUsM28AFgAB/Fn9COD5mfnLnVGcJEnSeLXZlrD65yn2ycyPDhs/JyIyMz0lKUmStI1GuybsS8A+I4yfBnyhPeVIkiTtHkYLYYdk5s+Gj8zM64Bnt68kSZKk8W+0EDZllGldO7oQSZKk3cloIWxZRLx6+MiIOAG4s30lSZIkjX+j/UTF+4DvRsQbGPoTFUcDr213YZIkSePZaD9RcTtwJPBTYDZwIPAT4O3Ae3dCbZIkSePWaC1hZOZ64OKImAecCnwUuAu4YifUJkmSNG6N9jthz6AKXqcCDwDfAiIzX7aTapMkSRq3RmsJ+x1wHfDazFwGEBHv3ylVSZIkjXOj/Xfk64B7gGsj4qsR8Qqq2xZJkiRpO412Yf53MvNNwGHAtVT/LblvRHw5Io7bWQVKkiSNR6O1hAGQmY9l5j9n5v8HzAJ+Bfz3tlcmSZI0jm0xhLXKzIcy88LMfEW7CpIkSdodbFUIkyRJ0o5hCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBbQ1hEXF8RNwWEcsi4uxR5vuvEZERMb+d9UiSJI0VbQthEdEALgBOAOYAp0bEnBHmmwK8F/hlu2qRJEkaa9rZErYAWJaZd2ZmL3ApcNII830c+BSwro21SJIkjSntDGEzgbtbhlfW4zaIiHnA/pn53dEWFBFnRsSiiFi0atWqHV+pJEnSTlbswvyI6AA+B/z1lubNzAszc35mzp8+fXr7i5MkSWqzdoawPwD7twzPqscNmgIcAfwkIpYDLwCu8uJ8SZK0O2hnCLsRODQiDoqIbuBNwFWDEzPzkczcJzNnZ+Zs4HrgxMxc1MaaJEmSxoS2hbDM7AfeBVwD3Apclpm3RMS5EXFiu9YrSZK0K+hs58Iz82rg6mHjPrKZeV/azlokSZLGEn8xX5IkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqoK0hLCKOj4jbImJZRJw9wvS/ioilEXFzRPwoIg5sZz2SJEljRdtCWEQ0gAuAE4A5wKkRMWfYbL8C5mfms4HLgU+3qx5JkqSxpJ0tYQuAZZl5Z2b2ApcCJ7XOkJnXZubj9eD1wKw21iNJkjRmtDOEzQTubhleWY/bnLcB3xtpQkScGRGLImLRqlWrdmCJkiRJZYyJC/Mj4i3AfOAzI03PzAszc35mzp8+ffrOLU6SJKkNOtu47D8A+7cMz6rHDRERxwJ/C7wkM9e3sR5JkqQxo50tYTcCh0bEQRHRDbwJuKp1hoh4LvB/gBMz8/421iJJkjSmtC2EZWY/8C7gGuBW4LLMvCUizo2IE+vZPgNMBv5vRCyJiKs2szhJkqRxpZ2nI8nMq4Grh437SEv/se1cvyRJ0lg1Ji7MlyRJ2t0YwiRJkgowhEmSJBVgCJMkSSrAECZJklSAIUySJKkAQ5gkSVIBhjBJkqQCDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmSZJUgCFMkiSpAEOYJElSAYYwSZKkAgxhkiRJBRjCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGMEmSpAIMYZIkSQUYwiRJkgowhEmSJBVgCJMkSSrAECZJklRAZ+kCdoS+vj5WrlzJunXrSpcyLvT09DBr1iy6urpKlyJJ0rg1LkLYypUrmTJlCrNnzyYiSpezS8tMVq9ezcqVKznooINKlyNJ0rg1Lk5Hrlu3jmnTphnAdoCIYNq0abYqSpLUZuMihAEGsB3IfSlJUvuNmxAmSZK0KzGE7QCrV69m7ty5zJ07lxkzZjBz5swNw729vU9qGaeffjq33XbbVq/7ta99Lcccc8wm4y+55BKOOOIIjjzySObNm8d5550HVNd8ffrTn+aZz3wmc+fO5XnPex7f/OY3t3q9kiRp+4yLC/NLmzZtGkuWLAHgnHPOYfLkyZx11llD5slMMpOOjpFz78UXX7zV633wwQe5+eab6enpYcWKFRxwwAEA/Ou//ivnn38+P/zhD5kxYwbr1q3jn/7pnwC44IILuPbaa1m0aBFTpkzhkUce4corr9zqdUuSpO0z7kLY+973byxZcu8OXebcuTP4/OeP3+rnLVu2jBNPPJHnPve5/OpXv+IHP/gBH/vYx7jpppt44okneOMb38hHPvIRAI455hjOP/98jjjiCPbZZx/e+c538r3vfY9JkyZx5ZVXsu+++26y/Msvv5yTTz6ZqVOncumll/LBD34QgE984hN87nOfY8aMGUD1kxNvf/vbN0z7xS9+wZQpUwCYOnUqp5122jbtF0mStO08Hdlmv/vd73j/+9/P0qVLmTlzJp/85CdZtGgRv/71r/nBD37A0qVLN3nOI488wkte8hJ+/etfc/TRR3PRRReNuOyFCxdy6qmncuqpp7Jw4cIN42+55RaOOuqoTeZ/8MEH6evr48ADD9xxGyhJkrbJuGsJ25YWq3Y6+OCDmT9//obhhQsX8rWvfY3+/n7++Mc/snTpUubMmTPkORMnTuSEE04A4KijjuK6667bZLl//NM3vuYAAA77SURBVOMfWbFiBUcffTQAzWaT3/3udxx22GFt3BpJkrSj2BLWZnvssceG/ttvv50vfOEL/PjHP+bmm2/m+OOPH/H3uLq7uzf0NxoN+vv7N5nnW9/6Fg888ACzZ89m9uzZrFixYkNr2Jw5c1i8ePEmz9l7773p6upixYoVO2LTJEnSdjCE7URr1qxhypQp7Lnnntxzzz1cc80127yshQsX8sMf/pDly5ezfPlybrjhhg0h7EMf+hBnnXUW9913HwDr16/na1/7GgBnn302f/EXf8HatWs31PSNb3xjO7dMkiRtrXF3OnIsmzdvHnPmzOGwww7jwAMP5EUvetE2LeeOO+7gnnvuGXKa89BDD6Wnp4fFixdz4oknsmrVKl7+8pcD1Y+vvuMd7wDg3e9+N4899hhHHXUU3d3ddHV1bbigX5Ik7TyRmaVr2Crz58/PRYsWDRl36623cvjhhxeqaHxyn0qStP0iYnFmzh9pmqcjJUmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqQBDmCRJUgGGsB1g9erVzJ07l7lz5zJjxgxmzpy5Ybi3t/dJL+eiiy7i3ns3f/Px3t5e9t57bz784Q8PGd/X18cHP/hBDjnkEObNm8cLX/jCDT8Eu2bNGt7xjndw8MEHc9RRR/Gyl72MG2+8cds2VJIk7TD+WOsOMG3aNJYsWQLAOeecw+TJkznrrLO2ejkXXXQR8+bNY8aMGSNOv+aaa5gzZw7f+ta3+Lu/+7sN4z/0oQ/x4IMPsnTpUrq7u7n33nv593//dwDOOOMMDj/8cJYtW0ZEcMcdd/D73/9+G7ZSkiTtSOMvhF37Prh/yY5d5r5z4WWf36anfv3rX+eCCy6gt7eXF77whZx//vk0m01OP/10lixZQmZy5plnst9++7FkyRLe+MY3MnHiRG644YYh95CE6lZFf/VXf8V5553HDTfcwIIFC1i7di2XXHIJy5cv3zD/jBkzeP3rX89tt93GkiVLuOyyy4gIoLqh+MEHH7x9+0OSJG238RfCxpDf/va3fPvb3+bnP/85nZ2dnHnmmVx66aUcfPDBPPDAA/zmN78B4OGHH2avvfbiS1/6Eueffz5z587dZFmPP/44P/nJTzacsly4cCELFizg9ttv56CDDmLy5MmbPOeWW27huc99Lh0dnnWWJGmsGX8hbBtbrNrhhz/8ITfeeOOGezw+8cQT7L///rzqVa/itttu4z3veQ+vec1rOO6447a4rKuuuopXvvKV9PT0cMopp3DUUUfx2c9+tt2bIEmS2mT8hbAxJDM544wz+PjHP77JtJtvvpnvfe97XHDBBVxxxRVceOGFoy5r4cKFXH/99cyePRuAVatW8dOf/pT58+dz11138eijj27SGvasZz2LJUuW0Gw2bQ2TJGmM8ZO5jY499lguu+wyHnjgAaD6L8oVK1awatUqMpNTTjmFc889l5tuugmAKVOmsHbt2k2W8/DDD3P99dezcuVKli9fzvLly/niF7/IwoULmTJlCqeddhrve9/76OvrA+D+++/n8ssv55nPfCZHHnkk5557LoM3ar/rrrv43ve+t5P2gCRJ2hxDWBsdeeSRfPSjH+XYY4/l2c9+Nscddxz33Xcfd999Ny9+8YuZO3cup59+Op/4xCcAOP3003n729++yU9bXHHFFbzyla+kq6trw7iTTz6Z73znO/T19fHJT36Svfbai8MPP5wjjzySE088kalTpwJw8cUXc/fdd3PIIYdwxBFHcMYZZ7Dvvvvu3B0hSZI2EYMtJLuK+fPn56JFi4aMu/XWWzn88MMLVTQ+uU8lSdp+EbE4M+ePNM2WMEmSpAIMYZIkSQWMmxC2q51WHcvcl5Iktd+4CGE9PT2sXr3a8LADZCarV6+mp6endCmSJI1r4+J3wmbNmsXKlStZtWpV6VLGhZ6eHmbNmlW6DEmSxrVxEcK6uro46KCDSpchSZL0pLX1dGREHB8Rt0XEsog4e4TpEyLiW/X0X0bE7HbWI0mSNFa0LYRFRAO4ADgBmAOcGhFzhs32NuChzDwEOA/4VLvqkSRJGkva2RK2AFiWmXdmZi9wKXDSsHlOAr5e918OvCIioo01SZIkjQntvCZsJnB3y/BK4Pmbmycz+yPiEWAa8EDrTBFxJnBmPfhoRNy2g2vdZ/g6Na55vHcvHu/dh8d697KrHO8DNzdhl7gwPzMvBC5s1/IjYtHmbimg8cfjvXvxeO8+PNa7l/FwvNt5OvIPwP4tw7PqcSPOExGdwFRgdRtrkiRJGhPaGcJuBA6NiIMioht4E3DVsHmuAt5a978e+HH6i6uSJGk30LbTkfU1Xu8CrgEawEWZeUtEnAssysyrgK8B34iIZcCDVEGthLad6tSY5PHevXi8dx8e693LLn+8w4YnSZKknW9c3DtSkiRpV2MIkyRJKmDch7CI2D8iro2IpRFxS0S8tx6/d0T8ICJur7tPqcdHRHyxvpXSzRExr+wWaFtERCMifhUR/1oPH1TfGmtZfaus7nq8t87axUXEXhFxeUT8LiJujYijfX2PXxHx/vq9/LcRsTAienx9jx8RcVFE3B8Rv20Zt9Wv54h4az3/7RHx1pHWNRaM+xAG9AN/nZlzgBcAf1nfPuls4EeZeSjwo3oYqtssHVo/zgS+vPNL1g7wXuDWluFPAefVt8h6iOqWWeCts8aDLwD/lpmHAc+hOu6+vsehiJgJvAeYn5lHUP3T15vw9T2eXAIcP2zcVr2eI2Jv4KNUPxC/APjoYHAba8Z9CMvMezLzprp/LdUb9EyG3jLp68DJdf9JwD9m5Xpgr4h46k4uW9shImYBrwH+oR4O4OVUt8aCTY+3t87aRUXEVODFVP9pTWb2ZubD+PoezzqBifVvS04C7sHX97iRmT+j+rWEVlv7en4V8IPMfDAzHwJ+wKbBbkwY9yGsVd0U/Vzgl8B+mXlPPeleYL+6f6TbLc3cSSVqx/g88EGgWQ9PAx7OzP56uPWYDrl1FjB46yztGg4CVgEX16ef/yEi9sDX97iUmX8A/hewgip8PQIsxtf3eLe1r+dd5nW+24SwiJgMXAG8LzPXtE6rfyDW3+oYByLitcD9mbm4dC3aKTqBecCXM/O5wGNsPFUB+PoeT+pTSidRhe+nAXswRls41B7j7fW8W4SwiOiiCmDfzMx/qUffN3gaou7eX49/Mrdb0tj1IuDEiFgOXEp1muILVM3Ugz9O3HpMvXXWrm0lsDIzf1kPX04Vynx9j0/HAndl5qrM7AP+heo17+t7fNva1/Mu8zof9yGsPv//NeDWzPxcy6TWWya9FbiyZfxp9X9dvAB4pKUZVGNcZn4oM2dl5myqC3Z/nJlvBq6lujUWbHq8vXXWLioz7wXujohn1qNeASzF1/d4tQJ4QURMqt/bB4+3r+/xbWtfz9cAx0XEU+rW0+PqcWPOuP/F/Ig4BrgO+A0brxH6G6rrwi4DDgD+E3hDZj5Yv7DPp2rifhw4PTMX7fTCtd0i4qXAWZn52oh4OlXL2N7Ar4C3ZOb6iOgBvkF1reCDwJsy885SNWvrRcRcqn/C6AbuBE6n+oLp63scioiPAW+k+s/3XwFvp7rex9f3OBARC4GXAvsA91H9l+N32MrXc0ScQfVZD/D3mXnxztyOJ2vchzBJkqSxaNyfjpQkSRqLDGGSJEkFGMIkSZIKMIRJkiQVYAiTJEkqwBAmaYeIiIyIz7YMnxUR5+ygZV8SEa/f8pzbvZ5TIuLWiLi23evaQh3LI2KfkjVIaj9DmKQdZT3wurEWHlp+Sf3JeBvwjsx8WbvqkaRBhjBJO0o/cCHw/uEThrdkRcSjdfelEfHTiLgyIu6MiE9GxJsj4oaI+E1EHNyymGMjYlFE/L6+RygR0YiIz0TEjRFxc0T8ectyr4uIq6h+UX14PafWy/9tRHyqHvcR4BjgaxHxmWHzPzUifhYRS+rn/Ek9/st1TbfUPyI6OP/yiPif9fyLImJeRFwTEXdExDtbavxZRHw3Im6LiK9ExCbvyRHxlnp/LImI/1Nvc6Pep7+tt2OTfS5p7Nuab4iStCUXADdHxKe34jnPAQ6n+kXzO4F/yMwFEfFe4N3A++r5ZgMLgIOBayPiEOA0qluVPC8iJgD/ERHfr+efBxyRmXe1riwingZ8CjgKeAj4fkScnJnnRsTLqe6yMPxX9P8UuCYz/z4iGsCkevzf1r/c3QB+FBHPzsyb62krMnNuRJwHXEJ1j8Me4LfAV+p5FgBzqH4F/N+A11Hd/3Kw1sOpfh3+RZnZFxH/G3gzcAswMzOPqOfb68nsaEljiy1hknaYzFwD/CPwnq142o2ZeU9mrgfuAAZD1G+ogtegyzKzmZm3U4W1w6juCXdaRCyhuhXZNODQev4bhgew2vOAn9Q3ge4Hvgm8eEs1AqfX17gdmZlr6/FviIibqG6V8yyqQDXoqpbt+GVmrs3MVcD6ltB0Q2bemZkDwEKqlrhWr6AKizfW2/gK4On19j89Ir4UEccDa7ZQv6QxyJYwSTva54GbgNZ7tfVTf+mrT7l1t0xb39LfbBluMvQ9avg91hII4N2ZOeTmvPV9Qx/btvI3lZk/i4gXA68BLomIz1Hdk/Ys4HmZ+VBEXELV0jWodTuGb+Pgdo20Ta0C+Hpmfmh4TRHxHOBVwDuBNwBnbO12SSrLljBJO1RmPkh1s923tYxeTtWiA3Ai0LUNiz4lIjrq68SeDtwGXAP8t4joAoiIZ0TEHltYzg3ASyJin/o04qnAT0d7QkQcCNyXmV+luln4PGBPqqD3SETsB5ywDdu0ICIOqoPpG4F/Hzb9R8DrI2Lfuo69I+LA+p8fOjLzCuDDdT2SdjG2hElqh88C72oZ/ipwZUT8murap21ppVpBFaD2BN6Zmesi4h+oTlneFBEBrAJOHm0hmXlPRJwNXEvV0vTdzLxyC+t+KfCBiOgDHgVOy8y7IuJXwO+Au4H/2IZtuhE4Hzikrufbw2pdGhEfprpurQPoA/4SeAK4uOVC/k1ayiSNfZE5vPVbktRu9SnTszLztaVrkVSGpyMlSZIKsCVMkiSpAFvCJEmSCjCESZIkFWAIkyRJKsAQJkmSVIAhTJIkqYD/B1APDZXieYuGAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "trn_score_acc = exponential_moving_average(score_trn_list[:,7], 0.1)\n",
    "tst_score_acc = exponential_moving_average(score_tst_list[:,7], 0.1)\n",
    "\n",
    "plt.title(\"ACC per samples\")\n",
    "plt.xlabel('Number of samples')\n",
    "plt.ylabel('ACC')\n",
    "plt.plot(train_size, trn_score_acc, label='Train ACC',color='navy')\n",
    "plt.plot(train_size, tst_score_acc, label='Test ACC',color=\"darkorange\")\n",
    "plt.legend(loc=\"best\")\n",
    "plt.ylim(0,1.05)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### F1-score F1 = 2TP / (2TP + FP + FN):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "trn_score_f1 = exponential_moving_average(score_trn_list[:,8], 0.1)\n",
    "tst_score_f1 = exponential_moving_average(score_tst_list[:,8], 0.1)\n",
    "\n",
    "plt.title(\"F1 per samples\")\n",
    "plt.xlabel('Number of samples')\n",
    "plt.ylabel('F1')\n",
    "plt.plot(train_size, trn_score_f1, label='Train F1',color='navy')\n",
    "plt.plot(train_size, tst_score_f1, label='Test F1',color=\"darkorange\")\n",
    "plt.legend(loc=\"best\")\n",
    "plt.ylim(0,1.05)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}