03_Make_Predictions_All_Features.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import glob\n",
    "import os\n",
    "import munge_help\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.utils.class_weight import compute_class_weight\n",
    "\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "\n",
    "#from sklearn.svm import SVC\n",
    "\n",
    "from sklearn.metrics import confusion_matrix,plot_roc_curve,plot_precision_recall_curve,plot_confusion_matrix, classification_report\n",
    "#from plot_help import plot_confusion_matrix\n",
    "\n",
    "\n",
    "from nltk.stem import WordNetLemmatizer \n",
    "\n",
    "import utils\n",
    "\n",
    "import xgboost as xgb\n",
    "\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#train data\n",
    "#X_train = utils.load_obj(path = os.path.join('data_processed', 'X_train.pkl'))\n",
    "\n",
    "y_train = utils.load_obj(path = os.path.join('data_processed', 'y_train.pkl'))\n",
    "\n",
    "X_train_num = utils.load_obj(path = os.path.join('data_processed', 'X_train_num.pkl'))\n",
    "\n",
    "X_train_text = utils.load_obj(path = os.path.join('data_processed', 'X_train_text.pkl'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test_num = utils.load_obj(path = os.path.join('data_processed', 'X_test_num.pkl'))\n",
    "\n",
    "X_test_text = utils.load_obj(path = os.path.join('data_processed', 'X_test_text.pkl'))\n",
    "\n",
    "y_test = utils.load_obj(path = os.path.join('data_processed', 'y_test.pkl'))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(102492, 111)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#explore numerical shape\n",
    "X_train_num.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(102492, 100)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#explore nlp shape\n",
    "X_train_text.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(102492, 211)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#combine horizontally\n",
    "X_train = np.concatenate((X_train_num, X_train_text.toarray()), axis=1)\n",
    "#check the shape with assertion\n",
    "assert X_train.shape[1] == (X_train_num.shape[1] + X_train_text.shape[1])\n",
    "#see the result\n",
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(102492, 211)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#combine horizontally\n",
    "X_test = np.concatenate((X_test_num, X_test_text.toarray()), axis=1)\n",
    "#check the shape with assertion\n",
    "assert X_test.shape[1] == (X_test_num.shape[1] + X_test_text.shape[1])\n",
    "#see the result\n",
    "X_train.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Logistic Reg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "#instantiate with default params\n",
    "lr = LogisticRegression(penalty='l2',\n",
    "                        dual=False, \n",
    "                        tol=0.0001, \n",
    "                        C=1.0, \n",
    "                        fit_intercept=True, \n",
    "                        intercept_scaling=1, \n",
    "                        class_weight='balanced', \n",
    "                        random_state=None, \n",
    "                        solver='lbfgs', \n",
    "                        max_iter=100, \n",
    "                        multi_class='auto', \n",
    "                        verbose=0, \n",
    "                        warm_start=False, \n",
    "                        n_jobs=-1, \n",
    "                        l1_ratio=None)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix in array form\n",
      "[[4475  860]\n",
      " [  14   46]]\n",
      "\n",
      "\n",
      "####################\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWwAAAEWCAYAAABCJq0eAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzt3Xm8VVXdx/HPl0FAkUHxUSbFcMoRh3CsTE3RfNLMAaMcUklTS9NKzErt0cxKy9IKE8EyFVOTzFSctUQZRARHFBUQB0AmB5TL7/ljr4uH67n3ngvncs++5/vutV/svfbaa69zbv7uumuvvZYiAjMzq3xtWroCZmZWGgdsM7OccMA2M8sJB2wzs5xwwDYzywkHbDOznHDAtmYnaYike1bx2mmS9i5zlSqepH9LOral62GVRR6HbYUkvQKcGBH3tsC9RwKzIuK81SynHzADeDclzQX+GBGXrE65Zi2tXUtXwKwZdYuIZZJ2AR6SNDEixpbzBpLaRcSycpZpVh93iVjJJJ0kabqk+ZLGSOpVcG5/Sc9LWijpKkkPSToxnTtO0qNpX5Iul/SWpEWSnpa0raShwBDgB5KWSPpnyv+KpP3SfltJ50p6SdJiSRMl9W2s3hExAZgGDCioby9Jt0h6W9IMSd8pONdJ0ihJ70h6VtIPJM0qOP+KpB9KmgK8K6ldI+UNlDQhfd43JV2W0jtK+qukeZIWSBovacN07sGC76+NpPMkvZq+t+skdU3n+kkKScdKek3SXEk/avIP13LBAdtKImkf4OfAkUBP4FXgxnSuB/B3YBiwPvA8sEc9Re0PfA7YAuiaypsXEcOB64FLI6JzRPxvkWu/BxwNHAR0Ab4JvFdC3XcDtgWmp+M2wD+Bp4DewL7AGZIOSJf8FOgHfAr4IvD1IsUeDXwJ6AYsb6S83wK/jYguQH9gdEo/Nn0Hfcm+t5OB94vc67i0fSHVqTPw+zp59gK2TPf+iaRPN/SdWD45YFuphgAjImJSRCwlC867p/7ig4BpEXFr6h64AnijnnI+AtYFtiJ7hvJsRMwpsQ4nAudFxPOReSoi5jWQf66k94HHgKuAf6T0zwAbRMSFEfFhRLwMXA0MTuePBC6OiHciYlb6PHVdEREzI+L9Esr7CNhMUo+IWBIR4wrS1wc2i4iaiJgYEYuK3GsIcFlEvBwRS8i++8GSCrs0L4iI9yPiKbJfHDs08L1YTjlgW6l6kbWqAUiBYx5Zi7IXMLPgXACz6haQzt1P1jq8EnhL0nBJXUqsQ1/gpSbUuQdZa/QsYG+gfUrfBOiVuiEWSFoAnAtsmM6v9Hnq7BdLa6y8E8j+ongudXscnNL/AtwN3CjpdUmXSmrPJ6303af9dgXlw8q/IN9Ln9taGQdsK9XrZIEJAEnrkLUOZwNzgD4F51R4XFdEXBEROwNbkwWy79eeaqQOM8m6FEqWWq6XAR8A3y4oZ0ZEdCvY1o2Ig9L5lT4P2S+KTxRdp171lhcRL0bE0cD/AL8A/i5pnYj4KCIuiIitybqQDgaOKXKvlb57YGNgGfBmE74KawUcsK2Y9umBWO3WDrgBOF7SAEkdgIuBxyPiFeBfwHaSDk15TwU2KlawpM9I2jW1JN8lC6TL0+k3yfpo6/Nn4GeSNk8PL7eXtH6Jn+kSsgeaHYEngMXpwWGn9DBzW0mfSXlHA8MkdZfUGzitkbIbLE/S1yVtEBHLgQXpmuWSviBpO0ltgUVkXSTLi5R/A3CmpE0ldSb77m/y6JTq44BtxdxJ9vCrdjs/jcv+MXALWQu0P6mPNiLmAkcAl5J1k2wNTACWFim7C1n/7jtkf9rPA36Zzl0DbJ26Ff5R5NrLyILpPWQB7hqgU4mf6V/pnidFRA1Za3YA2XjtuWS/DLqmvBeSdenMAO4le6Ba7LMAWSu+kfIGAdMkLSF7ADk49X1vlMpeBDwLPETWTVLXiJT+cCr/A+D0Ej+3tSJ+ccbKLo3CmAUMiYgHWro+q0vSKWRB9vMtXRerbm5hW1lIOkBSt9Rdci4gYFwjl1UkST0l7ZnGP29J9tDytpaul5nfdLRy2R34G7AW8AxwaPqzP4/WAv4EbErW53wj2bBAsxblLhEzs5xwl4iZWU64S2Q1rd+jbWy8ib/GPJk1qXdLV8GaaC4z5kbEBqtTxn77rx3z5tWUlHfypA/vjohBq3O/5uBIs5o23qQdD/zXASBPftDx4paugjXR1Rz9auO5GjZvXg0PPlbaf6vdOszosbr3aw4O2GZWJQTL27Z0JVaLA7aZVYcA1eT7sZ0DtplVBQFarpauxmpxwDaz6hCgYjO15IgDtplVDwdsM7McCFDO3xN0wDazquEuETOzPAhQTb6b2A7YZlY93MI2M6t82bA+t7DNzCpf4Ba2mVleeJSImVkeBCjnyxY7YJtZ9cj5gi0O2GZWNTwO28wsD/zQ0cwsP/zQ0cwsL3Lews73bN5mZiVSgGpU0lZSeVJbSU9KuiMdbyrpcUnTJd0kaa2U3iEdT0/n+xWUMSylPy/pgMbu6YBtZtVjeYlbab4LPFtw/Avg8ojYDHgHOCGlnwC8k9IvT/mQtDUwGNgGGARcJanBNcwcsM2sOtQ+dCxDwJbUB/gS8Od0LGAf4O8pyyjg0LR/SDomnd835T8EuDEilkbEDGA6MLCh+zpgm1n1iBK3xv0G+AEfh/f1gQURUftqziygdon23sBMgHR+Ycq/Ir3INUU5YJtZ1dBylbQBPSRNKNiGrihDOhh4KyImrun6e5SImVWHAEp8oAjMjYhd6jm3J/BlSQcBHYEuwG+BbpLapVZ0H2B2yj8b6AvMktQO6ArMK0ivVXhNUW5hm1n1KEMfdkQMi4g+EdGP7KHh/RExBHgAODxlOxa4Pe2PScek8/dHRKT0wWkUyabA5sATDd3bLWwzqw6l90+vqh8CN0r6P+BJ4JqUfg3wF0nTgflkQZ6ImCZpNPAMsAw4NSJqGrqBA7aZVQnB8pK7REoSEQ8CD6b9lykyyiMiPgCOqOf6i4CLSr2fA7aZVY8ob8Be0xywzaw6hGfrMzPLj9JHiVQkB2wzqw5B2fuw1zQHbDOrHu7DNjPLCfdhm5nlgdzCNjPLhYBwH7aZWU54lIiZWQ4E7hIxM8sNd4mYmeWBHzqameWDX5wxM8uP8ENHM7OccJeImVkOtIIuES8RZmZVIj10LGVrqBSpo6QnJD0laZqkC1L6SEkzJE1O24CULklXSJouaYqknQrKOlbSi2k7tr571nIL28yqR3la2EuBfSJiiaT2wKOS/p3OfT8i/l4n/4Fk6zVuDuwK/AHYVdJ6wE+BXcja/xMljYmId+q7sVvYZlYVIkrfGi4nIiKWpMP2aWvoqkOA69J148hWV+8JHACMjYj5KUiPBQY1dG8HbDOrHjVtStugh6QJBdvQwmIktZU0GXiLLOg+nk5dlLo9LpfUIaX1BmYWXD4rpdWXXi93iZhZdQiI0keJzI2IXeotKlvdfICkbsBtkrYFhgFvAGsBw8lWUb9w9Sq9MrewzaxKpFXTS9lKFBELgAeAQRExJ3V7LAWu5eMV1GcDfQsu65PS6kuvl1vYVaimRhy891fYsNe7jLzp7hXpP/nBHoy+fkuem30tABcM253HHukJwPvvt2Pe252Y+tooAPqtdyJbbT0fgF593mXEjXdjze+z33mUgcdPgIA5Uzdi9ElfZdnSdgy6YCzbf/Vplte04bHhu/KfK/cAgkMuu4OtBj3PR++txU0nfpXZkxv8i7v1K8M4bEkbAB9FxAJJnYAvAr+Q1DMi5kgScCgwNV0yBjhN0o1kDx0Xpnx3AxdL6p7y7U/WSq9XswVsSQFcFhFnpeOzgc4RcX46Hgp8L2VfBHwvIh5txvocB+wSEac1kOdk4L2IuC7lvyciXm+uOrWUEX/Yls22XMDixe1XpD31ZA8WLuiwUr6f/vyxFfvX/mkbpk3pseK4Y6ca7nr01uavrK3QpddC9jr1MX65wxks+6A9X7/+bww4cgoIuvVZyC+3O5OINqyzQfY8bKtBL9Bjs3n8Yuuz2HjgTA773e387rPfbuFP0bKa0CXSkJ7AKEltyXopRkfEHZLuT8FcwGTg5JT/TuAgYDrwHnB8VpeYL+lnwPiU78KImN/QjZuzS2QpcJikHnVPSDoY+BawV0RsRfbB/iZpo2asT6Mi4o8RcV06PA7o1YLVaRZzZq/DffdszOBvPLciraZGXPzj3Tj3wnH1Xjfmlv58+fDpa6KK1oA2bZfTvtNHtGlbQ/u1P2LRnC7sPvRxxl68DxHZf87vvt0ZgG3+9xkm/nVHQLz2xMZ07PYB6260qAVr38KCbImwUraGiomYEhE7RsT2EbFtRFyY0veJiO1S2tdrR5KkbpJTI6J/Oj+hoKwREbFZ2q5t7CM0Z8BeRtbxfmaRcz8kG684FyAiJgGjgFPrZpTUX9JdkiZKekTSVin9dknHpP1vSbo+7T8o6bdp4PpUSQOLlNkv/TacIuk+SRun9PMlnS3pcLKxkdencjqV4wupBOcP251zL3ycNm0+HoU0cvg2fPHAV9lwo/eLXjPrtc689moX9vzcx39sLP2gLV/a+yscst8h3H3HJs1eb4NFr3flod/sxY+mX8qPX/05HyzsyAv3bs76n5rHDodP4Tv/vZITxoykx2ZzAejSaxELZnVdcf3C2V3o2quKAzYQNW1K2ipVc9fsSmCIpK510rcBJtZJm5DS6xoOnB4ROwNnA1el9KHATyR9FjgLOL3gmrUjYgDwbWBEkTJ/B4yKiO2B64ErCk+mge8TgCERMSAiVopkkobWDveZ+3ZNsc9dke69a2N6bPA+2w+YuyLtjTlr86/bP8Vx35pa73Vjbu3Pl778Mm3bfhzkH3v6b/zrwdu44s/3c8GwPXhlxrrNWneDTt3eZ5uDn+XnW57Nz/oNY611PmSno5+kXYcali1txxV7nMrj1+zCEX+6paWrWplKfcuxgucbadaHjhGxSNJ1wHeA4s23BkjqDOwB3Jz14wPQIZX9pqSfkD2h/Uqdvp8bUp6HJXVJQ28K7Q4clvb/AlzalHpFxHCyXyTsuHOHRobZV44Jj2/I2H9vwgP3bMzSpW1ZvHgt9tvtCDp0qOFzOw4G4P332vHZHY/ikSdvWnHdP2/pz89+9Z+Vytqo13sAbNJvMbvt9TrTpvSg36aL19yHqUKb7zOd+a905925WZfH1H9swya7v8bC2V14+h9ZW2fq7dtw5NVZwF70ehe69Vm44vquvRex8PUua77iFaRMfdgtZk2MEvkNMIlsmEutZ4CdgfsL0nYGptW5tg2wILWWi9kOmMcn+5rrBtHcBNXmdM5Px3POT7PnG4890pM//X77lUaJAGzV+/iVgvX0F7qycEEHdh745oq0BQvWolOnZXTosJz58zow4fGNOPm7T62ZD1HF3pnZjY13nUn7Th/y0fvt2ewLLzFrUm8+WNSBzT7/MuNfWY9PfW4Gc1/MHhtNu+PT7HnKOCaP3p6NB87kg4UdWfxGdQfsvE/+1OwBOz0JHQ2cwMfdE5eSDYMZFBHz0iQpx5ENeSm8dlGaTOWIiLg5DZfZPiKeSn3TBwI7Ag9JuiciZqRLjwIekLQX2RCahQUtdID/AoPJWtdDgEeKVH0xUPV/54+5ZTP+96svUfj1TX++O8PO/CxtFCwP8e0zJrPFVgtarpJVYub4vjx967ac8fjvWb6sDbMn92LcnwfSvtNHfG3UaD77nf/w4ZK1uPnk7I/H5/69JZ8e9DznPPtrPnyvPaNP+moLf4IKkPMWtqKxF+dXtWBpSUR0TvsbAjOASwuG9Z0CnEHW+l0MnBURDxcpZ1OyyVJ6kr2zfyPwC+AJ4PiImCTpy2QPN/ch6yKZDHw+5f9mRDxROKxP0iZkLf4ewNupnNcknQ8siYhfSfoqcDFZV87udfuxa+24c4d44L9VPrY1Z37Q8eKWroI10dUcPbGhNw9LsVPfLvGfM3crKe/aZ41d7fs1h2ZrYdcG67T/JrB2nfN/IAvEjZUzg+ITouxQkGcM2eB0Ukv6rxFxRp1yRgIj0/6rZMG97r3OL9i/BfDTG7NWQ+7DNjPLDQfsyhIRe7d0HcysAgWEHzqameWEW9hmZvngPmwzszwIETUO2GZmFS9wC9vMLD/80NHMLAeatkRYRarceQTNzMqtDLP1Seoo6QlJT0maJumClL6ppMclTZd0k6S1UnqHdDw9ne9XUNawlP68pAMaq74DtplViexNx1K2RiwF9omIHYABwCBJu5FNmXF5RGwGvEM2fxLp33dS+uUpH5K2JpvTaBuyt7mvSqvY1MsB28yqQ0DUqKStwWIyS9Jh+7QF2XQXf0/po8jWdQQ4JB2Tzu+bJrI7BLgxIpamKTim8/HCvUU5YJtZVagdJVJiC7tH7SIlaRtaWJaktpImA28BY4GXyKaCXpayzAJqZ4XrDcwESOcXAusXphe5pig/dDSz6lH6Q8e5Dc3WFxE1wIC0OMptwFZlqF2j3MI2syohYnlpW6kiYgHZlM67A90k1TaC+wCz0/5soC9AOt+VbOGVFelFrinKAdvMqkM0qUukXpI2qF12MC3Q/UXgWbLAfXjKdixwe9ofk45J5++PbCGCMcDgNIpkU2Bzsnn+6+UuETOrGrG8LG3UnsCoNKKjDTA6Iu6Q9Axwo6T/A54Erkn5rwH+Imk6MJ9sZAgRMS2txvUMsAw4NXW11MsB28yqRiwvQxkRU8iWJqyb/jJFRnlExAfAEfWUdRFwUan3dsA2s+oQeHpVM7M8CC8RZmaWHw7YZmZ54YBtZpYDActr8j2S2QHbzKpHtHQFVo8DtplVCT90NDPLhVa9RJikLg1dGBGLyl8dM7NmEjRpnpBK1FALexrZL6XCT1h7HMDGzVgvM7OyK9Or6S2m3oAdEX3rO2dmlj/578Mu6deNpMGSzk37fSTt3LzVMjMrs4AocatUjQZsSb8HvgB8IyW9B/yxOStlZlZuTVxxpiKVMkpkj4jYSdKTABExv3Y1YDOzXGnFDx1rfSSpDWnIuaT1gTJMUmhmtmZVcuu5FKX0YV8J3AJsIOkC4FHSMu1mZrkRYvnyNiVtDZHUV9IDkp6RNE3Sd1P6+ZJmS5qctoMKrhkmabqk5yUdUJA+KKVNl3ROYx+h0RZ2RFwnaSKwX0o6IiKmNnadmVmlKVMLexlwVkRMkrQuMFHS2HTu8oj4VWFmSVuTrTKzDdALuFfSFun0lWRLjM0CxksaExHP1HfjUt90bAt8RNYtku+BjGZWvcoQsCNiDjAn7S+W9CzQu4FLDgFujIilwIy0VFjtyjTT00o1SLox5a03YJcySuRHwA1kvxn6AH+TNKzRT2VmVkEisiXCStlKJakf2XJhj6ek0yRNkTRCUveU1huYWXDZrJRWX3q9SmktHwN8JiLOi4gfkf1mOK6E68zMKkoThvX1kDShYBtatyxJncme752Rpur4A9AfGEDWAv91uetfSpfInDr52qU0M7NcaUIf9tyI2KW+k5LakwXr6yPi1qzseLPg/NXAHelwNlD45niflEYD6UU1NPnT5WR91vOBaZLuTsf7A+MbKtTMrPKo0REgJZUiCbgGeDYiLitI75n6twG+AtQOzhhD1pV8GVnX8ubAE2TzMm0uaVOyQD0Y+FpD926ohV17s2nAvwrSx5XyoczMKkr5Vk3fk+zN76clTU5p5wJHSxqQ7vQK8C2AiJgmaTTZw8RlwKkRUQMg6TTgbrKBHSMiYlpDN25o8qdrVucTmZlVknLNhx0Rj7LyLKa17mzgmouAi4qk39nQdXU12octqX+60dZAx4IbbVHvRWZmFaga3nQcCVxL9hvlQGA0cFMz1snMrPyaYVjfmlZKwF47Iu4GiIiXIuI8ssBtZpYj5Xk1vSWVMqxvaZr86SVJJ5M9zVy3eatlZlZerXpNxwJnAusA3yHry+4KfLM5K2Vm1hxafcCOiNpXLhfz8SIGZmb5Eq04YEu6jTQHdjERcViz1MjMrFlU9moypWiohf37NVaLHJvy5Ab07XpSS1fDmqBdh1daugrWVEvLVE5rXXEmIu5bkxUxM2tOEVT0CJBSlDoftplZ7lXyiuilcMA2s6rRmvuwVyKpQ1oxwcwsh/L/0LGUFWcGSnoaeDEd7yDpd81eMzOzMmvCAgYVqZQe+CuAg4F5ABHxFPCF5qyUmVm5ReQ/YJfSJdImIl7N5uxeoaaZ6mNm1myW17T+USIzJQ0EQlJb4HTgheatlplZuVV267kUpfy6OQX4HrAx8CawW0ozM8uPMnWJSOor6QFJz0iaJum7KX09SWMlvZj+7Z7SJekKSdPTiuo7FZR1bMr/oqRjG/sIpcwl8hbZWmNmZrlVxtn6lgFnRcQkSesCEyWNBY4D7ouISySdA5wD/JBsOurN07Yr2erqu0paD/gpsEuq3kRJYyLinfpuXMqKM1dTZE6RiPjEsu9mZpWsTEuEzQHmpP3Fkp4FegOHAHunbKOAB8kC9iHAdRERwDhJ3ST1THnHRsR8gBT0BwE31HfvUvqw7y3Y70i2GvDMEj+bmVmFaNKq6T0kTSg4Hh4Rwz9RotQP2BF4HNiwYNX0N4AN035vVo6Zs1Jafen1KqVLZKXlwCT9BXi0sevMzCpKQJQ++dPciNiloQySOgO3AGdExKLCkXQREZLK/iL8qoxx2ZSPf3OYmeVCbR92OcZhS2pPFqyvj4hbU/KbqauD9O9bKX020Lfg8j4prb70epXypuM7kuanbQEwFhjW6CcyM6sw2cszjW8NUdaUvgZ4NiIuKzg1Bqgd6XEscHtB+jFptMhuwMLUdXI3sL+k7mlEyf4prV4Ndomkiu3Ax1F/eeo4NzPLneXlGSWyJ9nqW09LmpzSzgUuAUZLOgF4FTgynbsTOAiYDrwHHA8QEfMl/QwYn/JdWPsAsj4NBuzUD3NnRGzb9M9kZlZByrREWEQ8CtRX0L5F8gdwaj1ljQBGlHrvUkaJTJa0Y0Q8WWqhZmaVJpo2SqQiNbSmY7uIWEY2ZGW8pJeAd8l+s0RE7FTftWZmlSjvr6Y31MJ+AtgJ+PIaqouZWfNp2rC+itRQwBZARLy0hupiZtasWnMLewNJ36vvZJ3hLGZmFS1awWx9DQXstkBn6n8aamaWK605YM+JiAvXWE3MzJpTQE1rHSWCW9Zm1oqUcXrVFtNQwP7EAHAzszyL5S1dg9VTb8Bu7BVJM7N8ad0PHc3MWo8o21wiLcYB28yqQkDrfTXdzKy1cZeImVkuyF0iZmZ5UMriBJUu3x06ZmZNEMtV0tYYSSMkvSVpakHa+ZJmS5qctoMKzg2TNF3S85IOKEgflNKmSzqnsfs6YJtZ1SjXmo7ASGBQkfTLI2JA2u4EkLQ1MBjYJl1zlaS2ktoCVwIHAlsDR6e89XKXiJlVhQioKdP0qhHxsKR+JWY/BLgxIpYCMyRNBwamc9Mj4mUASTemvM/UV5Bb2GZWNZrQwu4haULBNrTEW5wmaUrqMume0noDMwvyzEpp9aXXyy1sM6sSTRolMjcidmniDf4A/IxsyPfPgF8D32xiGQ1ywDazqpBN/tSM5Ue8Wbsv6WrgjnQ4G+hbkLVPSqOB9KLcJWJmVaOMDx0/QVLPgsOvALUjSMYAgyV1kLQpsDnZEozjgc0lbSppLbIHk2Mauodb2GZWHQJqasrz0FHSDcDeZH3ds4CfAntLGpDdiVeAbwFExDRJo8keJi4DTo2ImlTOacDdZAvGjIiIaQ3d1wHbzKpCOefDjoijiyRf00D+i4CLiqTfCdxZ6n0dsM2sSvjVdDOzfGgFr6Y7YJtZVQg8H7aZWW64hW1mlhPlejW9pThgm1lVaA3Tqzpgm1nVcB+2mVlO5L2F7VfTDYBfXfkQk1/6C/eO+/snzg09bQqzFl1N9/U+aIGaWUPatFnOvx65hWtG35VSgrN//AT3T7qJe8eP5riTpzZ4fbWp7RZpbKtUazxgS6opWJFhcimrLNRTzkhJhzeS58+1E4JLOncV7nGcpN+vSv3y5ubrt+Drhx34ifSevZfwuX1nMeu1zi1QK2vM8adMZfoL3VYcHzHkBXr2eZd9dz6S/T5zJP/8e/8WrF1lqR3WV8pWqVqihf1+wYoMAyLikua6UUScGBG1k4E3OWBXk8f/25MF73T4RPr5Px/HRT/etaJbHdVqo15L2OeA17hx1FYr0oac+AxX/GKnFa9gz5vbqaWqV3kCakrcKlVFdIlI6prWNdsyHd8g6aS0v0TS5ZKmSbpP0gZFrt9X0pOSnk4Th3dI6Q9K2kXSJUCn1KK/Pp37uqQnUtqf0nI9SDpe0guSngD2XFPfQSXa/6BXeGPO2jw7df2WrooV8ZNLHuPnP9l1pTUIN9l0EQcf9hJjHryVkbf8m379F7ZgDStLoJK3StUSAbs2cNZuR0XEQuA0YKSkwUD3iLg65V8HmBAR2wAPkc2KtYKkjmTrqx0VEduRPUg9pTBPRJzDxy37IZI+DRwF7BkRA4AaYEiaHvECskC9F9k6a58gaWjtShQR75bhK6k8HTst4/SzJ/Ori5o6h7utCfsMepV5czsxdfLK7Ze11qph6Qft+PLeh3HDyK249MqHWqiGlWl5lLZVqpYYJfJ+CpIriYixko4gW5Ryh4JTy4Gb0v5fgVvrXLolMCMiXkjHo4BTgd80UId9gZ2B8ZIAOgFvAbsCD0bE2wCSbgK2KFLX4cBwgLZt+lTwj3fV9dt0EX03Wcw9/7kFgJ693+WuR27l4C8cyttvrd3CtbNddn2T/Q58lS988TU6dKyh87ofcvnV9/PG6+tw1z/7AXD3P/vxy6sebNF6Vpq8/8daMcP6JLUBPg28B3QnW9+smHJ85wJGRcSwOnU4tAxltwrPPbMeA/p/Y8XxY0/fwEGf/wrvzO/YgrWyWpdeMJBLL8jWcd1tr9c56TtTOPOkffjh+Y+z+2df5+ZXu7DbXnOY8VK3RkqqHtlDx5auxeqpiD7s5EzgWeBrwLWS2qf0NkDtaJCvAY/Wue55oJ+kzdLxN8i6Tur6qKDM+4DDJf0PgKT1JG0CPA58XtL6Ke8RZfhcufD7Efdz+72303/zBYx/9m8M/sZzLV2jLBiWAAAMU0lEQVQlWwV/uHwABx4yg7seu5kfnP8E55z2uZauUkXJ+0PHlmhhd5I0ueD4LuBa4ERgYEQslvQwcB5Zf/W7wEBJ55F1WxxVWFhEfCDpeOBmSe3Ilt35Y5H7DgemSJqU+rHPA+5JLfuPyFaBGCfpfOAxYAEwuUg5rdJp39ynwfO7b1dsvnarBOMe7cW4R3sBsGhhB755xCeHZ1qmXLFY0gjgYOCtiNg2pa1H1n3bj2zFmSMj4h1l/a6/BQ4i60E4LiImpWuOJYt1AP8XEaMavG9U+HgtSUsiomIHAbdt0yfWXuvUlq6GNUG7Ch4FYMUtWDps4iqsYr6SjdQ/jtHFJeX9ZQxu8H6SPgcsAa4rCNiXAvMj4pL0fkn3iPihpIOA08kC9q7AbyNi1xTgJwC7kP0umQjsHBHv1HffSuoSMTNrVlHi1mg5EQ8D8+skH0I26IH076EF6ddFZhzQLY1IOwAYGxHzU5AeCwxq6L4V89CxPpXcujazfFleetYekiYUHA9Po8MasmFEzEn7bwAbpv3ewMyCfLNSWn3p9ar4gG1mVg7ZIrwlZ5+7Ol0wERGSyt7f7C4RM6saNSVuq+jN1NVB+vetlD4b6FuQr09Kqy+9Xg7YZlYVgqxLpJRtFY0Bjk37xwK3F6Qfo8xuwMLUdXI3sL+k7pK6A/untHq5S8TMqsZqBOOVSLoB2Jusr3sW2RDkS4DRkk4AXgWOTNnvJBshMp1sWN/xABExX9LPyIYiA1wYEXUfZK7EAdvMqka5OpUjor4XE/YtkjfIpssoVs4IYESp93XANrOqUNslkmcO2GZWJYLI+fRPDthmVjVWYwRIRXDANrOq4C4RM7MciVLfZanQnhMHbDOrGm5hm5nlgLtEzMxypKZS+zpK5IBtZlUhmzrVAdvMLBfcJWJmlhNR6mJDFdoQd8A2s6qQPXSs0EhcIgdsM6sa7hIxM8uBIDxKxMwsL/LeJeIVZ8ysaoRK2xoj6RVJT0uaXLtYr6T1JI2V9GL6t3tKl6QrJE2XNEXSTqtafwdsM6sKtQ8dS9lK9IWIGFCwWO85wH0RsTlwXzoGOBDYPG1DgT+s6mdwwDazqhEl/m8VHQKMSvujgEML0q+LzDigW+1ivU3lgG1mVaMJi/D2kDShYBtap6gA7pE0seDchmlxXYA3gA3Tfm9gZsG1s1Jak/mho5lVhSaOEplb0NVRzF4RMVvS/wBjJT230r0iQip1LtfSOWCbWdVYXqYYGhGz079vSboNGAi8KalnRMxJXR5vpeyzgb4Fl/dJaU3mLhEzqwrleugoaR1J69buA/sDU4ExwLEp27HA7Wl/DHBMGi2yG7CwoOukSdzCNrOqUaY+ig2B2yRBFkP/FhF3SRoPjJZ0AvAqcGTKfydwEDAdeA84flVv7IBtZlWjHC/ORMTLwA5F0ucB+xZJD+DU1b4xDthmViUCWJbzNx0dsM2sSqzWGOuK4IBtZlXB06uameWFyjesr6U4YJtZVfCq6WZmOeIuETOzHMheTc93G9sB28yqhlvYZmY54YBtZpYDHtZnZpYjy0tY/quSOWCbWVVwC9vMLCeC4COPEjEzywe3sM3McsIB28wsB4KgRu4SMTOreAFNWYS3Ijlgm1lVCODDnLewla1eY6tK0ttk67e1Rj2AuS1dCWuS1voz2yQiNlidAiTdRfb9lGJuRAxanfs1Bwdsq5ekCRGxS0vXw0rnn1nr1qalK2BmZqVxwDYzywkHbGvI8JaugDWZf2atmPuwzcxywi1sM7OccMA2M8sJB+yckhSSfl1wfLak8wuOh0p6Lm1PSNqrmetznKTfN5LnZEnHFOTv1Zx1qlSSaiRNLtjOWcVyRko6vJE8f5a0ddo/dxXu0ejP1dYcv+mYX0uBwyT9PCJWelFC0sHAt4C9ImKupJ2Af0gaGBFvtERlASLijwWHxwFTgddbpjYt6v2IGLAmbhQRJxYcngtcvCbua83DLez8WkY2IuDMIud+CHy/NpBHxCRgFHBq3YyS+ku6S9JESY9I2iql317QGv6WpOvT/oOSfptahlMlDSxSZj9J90uaIuk+SRun9PPTXwKHA7sA16dyOpXjC8kzSV0lPS9py3R8g6ST0v4SSZdLmpa+z0+88SdpX0lPSnpa0ghJHVL6g5J2kXQJ0Cl937U/y6+nv74mS/qTpLYp/XhJL0h6AthzTX0H1jgH7Hy7EhgiqWud9G2AiXXSJqT0uoYDp0fEzsDZwFUpfSjwE0mfBc4CTi+4Zu3UQvw2MKJImb8DRkXE9sD1wBWFJyPi76k+QyJiQES83/DHbHVqA2ftdlRELAROA0ZKGgx0j4irU/51gAkRsQ3wEPDTwsIkdQRGAkdFxHZkfzmfUpgnIs4htewjYoikTwNHAXumn2UN2f+XegIXkAXqvYCtm+UbsFXiLpEci4hFkq4DvgM0OehJ6gzsAdwsrVjsrkMq+01JPwEeAL4SEfMLLr0h5XlYUhdJ3eoUvTtwWNr/C3BpU+vWyhXtEomIsZKOIPtFvEPBqeXATWn/r8CtdS7dEpgRES+k49q/pn7TQB32BXYGxqeffSfgLWBX4MGIeBtA0k3AFqV/NGtODtj59xtgEnBtQdozZP8x3l+QtjMwrc61bYAFDfSnbgfMA+o+HKw7eN+D+ctAUhvg08B7QHdgVj1Zy/F9i+yvoGF16nBoGcq2ZuIukZxLLd/RwAkFyZcCv5C0PoCkAWQP+a6qc+0iYEZq1aHMDml/IHAgsCNwtqRNCy49KuXZC1iY/pwv9F9gcNofAjxSpOqLgXWb9GFbvzOBZ4GvAddKap/S2wC1o0G+Bjxa57rngX6SNkvH3yDrOqnro4Iy7wMOl/Q/AJLWk7QJ8DjweUnrp7xHlOFzWZm4hd06/Jqs/xOAiBgjqTfwX0lBFhy/HhFzilw7BPiDpPOA9sCNkp4DrgaOj4jXJZ0FjJC0T7rmA0lPpvzfLFLm6WQB5/vA28DxRfKMBP4o6X1g9yrrx+4kaXLB8V1kfyGdCAyMiMWSHgbOI+uvfhcYmH5Gb5F+YdaKiA8kHU/WtdUOGA8UjsipNRyYImlS6sc+D7gntew/Ak6NiHHKhoc+BiwAJhcpx1qIX023JpH0IHB2RExo6bpUC0lLIqJzS9fDWp67RMzMcsItbDOznHAL28wsJxywzcxywgHbzCwnHLCt2enj2emmSrpZ0tqrUdbeku5I+19WAzPdSeom6durcI/zJZ1danqdPI3OoFcnfz9JU5taR6tODti2JtTOYbEt8CFwcuHJ9MJOk/+/GBFjIuKSBrJ0I5vvxKxVcMC2Ne0RYLPUsnw+zYUyFegraX9Jj0malFrinQEkDVI2r/ckPp6jZKW5miVtKOk2SU+lbQ/gEqB/at3/MuX7vqTxymYSvKCgrB+lGeoeJZubo0GSTkrlPCXpljp/NewnaUIq7+CUv62kXxbc+1ur+0Va9XHAtjUmvYV3IPB0StocuCrNQvcu2Zt9+0XETmSz+X0vzUR3NfC/ZPOhbFRP8VcAD0XEDsBOZPOmnAO8lFr335e0f7rnQGAAsLOkz0namexV+gHAQcBnSvg4t0bEZ9L9nmXlqQH6pXt8iextzo7p/MKI+Ewq/6Q6r/ubNcqvptuaUPgq9iPANWQTSr0aEeNS+m5kU3n+J80etxbZ69Fbkc1E9yKApL+STf1a1z7AMQARUQMslNS9Tp790/ZkOu5MFsDXBW6LiPfSPcaU8Jm2lfR/ZN0unYG7C86NjojlwIuSXk6fYX9g+4L+7a7p3i9gViIHbFsTPjGdaArK7xYmAWMj4ug6+cq5MouAn0fEn+rc44xVKGskcGhEPCXpOGDvgnPFZjMU2bzjhYEdSf1W4d5WpdwlYpViHLBn7YxzktaRtAXwHNlMdP1TvqPruf4+0qT9qb+4K5+cEfBu4JsFfeO902x1DwOHSuokaV2y7pfGrAvMSTPaDalz7ghJbVKdP0U2m97dwCm1s+VJ2kLSOiXcx2wFt7CtIkTE26mleoPS8lbAeRHxgqShwL8kvUfWpVJsWtbvAsMlnUC2esopEfGYpP+kYXP/Tv3YnwYeSy38JWSzGE5SNlH/U2Sz4Y0voco/JpuK9O30b2GdXgOeALoAJ6fZ9P5M1rc9SdnN3wY897Q1iecSMTPLCXeJmJnlhAO2mVlOOGCbmeWEA7aZWU44YJuZ5YQDtplZTjhgm5nlxP8Dr0oiAOrIGQ0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "####################\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  NO exploit       1.00      0.84      0.91      5335\n",
      "   Exploited       0.05      0.77      0.10        60\n",
      "\n",
      "    accuracy                           0.84      5395\n",
      "   macro avg       0.52      0.80      0.50      5395\n",
      "weighted avg       0.99      0.84      0.90      5395\n",
      "\n"
     ]
    }
   ],
   "source": [
    "lr.fit(X_train, y_train)\n",
    "\n",
    "y_lr = lr.predict(X_test)\n",
    "\n",
    "utils.metric_evaluation(lr, X_train, X_test, y_test, 'Logistic Regression')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Naive Bayes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "nb = GaussianNB(priors=None, \n",
    "                var_smoothing=1e-09)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix in array form\n",
      "[[ 911 4424]\n",
      " [   3   57]]\n",
      "\n",
      "\n",
      "####################\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "####################\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  NO exploit       1.00      0.17      0.29      5335\n",
      "   Exploited       0.01      0.95      0.03        60\n",
      "\n",
      "    accuracy                           0.18      5395\n",
      "   macro avg       0.50      0.56      0.16      5395\n",
      "weighted avg       0.99      0.18      0.29      5395\n",
      "\n"
     ]
    }
   ],
   "source": [
    "nb.fit(X_train, y_train)\n",
    "\n",
    "y_nb = nb.predict(X_test)\n",
    "\n",
    "utils.metric_evaluation(nb, X_train, X_test, y_test, 'Logistic Regression')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "knn = KNeighborsClassifier(n_neighbors=5,\n",
    "                           weights='uniform', \n",
    "                           algorithm='auto', \n",
    "                           leaf_size=30, \n",
    "                           p=2, \n",
    "                           metric='minkowski', \n",
    "                           metric_params=None, \n",
    "                           n_jobs=-1)\n",
    "\n",
    "knn.fit(X_train, y_train)\n",
    "y_knn = knn.predict(X_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix in array form\n",
      "[[5323   12]\n",
      " [  55    5]]\n",
      "\n",
      "\n",
      "####################\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "####################\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  NO exploit       0.99      1.00      0.99      5335\n",
      "   Exploited       0.29      0.08      0.13        60\n",
      "\n",
      "    accuracy                           0.99      5395\n",
      "   macro avg       0.64      0.54      0.56      5395\n",
      "weighted avg       0.98      0.99      0.98      5395\n",
      "\n"
     ]
    }
   ],
   "source": [
    "utils.metric_evaluation(knn, X_train, X_test, y_test, 'KNN')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "rf = RandomForestClassifier(n_estimators=100,\n",
    "                            criterion='gini', \n",
    "                            max_depth=None, \n",
    "                            min_samples_split=2, \n",
    "                            min_samples_leaf=1, \n",
    "                            min_weight_fraction_leaf=0.0, \n",
    "                            max_features='sqrt', \n",
    "                            max_leaf_nodes=None, \n",
    "                            min_impurity_decrease=0.0, \n",
    "                            min_impurity_split=None, \n",
    "                            bootstrap=True, \n",
    "                            oob_score=False, \n",
    "                            n_jobs=-1, \n",
    "                            random_state=None, \n",
    "                            verbose=0, \n",
    "                            warm_start=False, \n",
    "                            class_weight='balanced', \n",
    "                            ccp_alpha=0.0, \n",
    "                            max_samples=None)\n",
    "rf.fit(X_train, y_train)\n",
    "y_rf = rf.predict(X_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix in array form\n",
      "[[5311   24]\n",
      " [  51    9]]\n",
      "\n",
      "\n",
      "####################\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "####################\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  NO exploit       0.99      1.00      0.99      5335\n",
      "   Exploited       0.27      0.15      0.19        60\n",
      "\n",
      "    accuracy                           0.99      5395\n",
      "   macro avg       0.63      0.57      0.59      5395\n",
      "weighted avg       0.98      0.99      0.98      5395\n",
      "\n"
     ]
    }
   ],
   "source": [
    "utils.metric_evaluation(rf, X_train, X_test, y_test, title = 'Random Forest')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### GBTree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "gb = GradientBoostingClassifier(loss='deviance', \n",
    "                                    learning_rate=0.1, \n",
    "                                    n_estimators=100, \n",
    "                                    subsample=1.0, \n",
    "                                    criterion='friedman_mse', \n",
    "                                    min_samples_split=2, \n",
    "                                    min_samples_leaf=1, \n",
    "                                    min_weight_fraction_leaf=0.0, \n",
    "                                    max_depth=3, \n",
    "                                    min_impurity_decrease=0.0, \n",
    "                                    min_impurity_split=None, \n",
    "                                    init=None, \n",
    "                                    random_state=None, \n",
    "                                    max_features='sqrt', \n",
    "                                    verbose=0, \n",
    "                                    max_leaf_nodes=None, \n",
    "                                    warm_start=False, \n",
    "                                    presort='deprecated', \n",
    "                                    validation_fraction=0.1, \n",
    "                                    n_iter_no_change=None, \n",
    "                                    tol=0.0001, \n",
    "                                    ccp_alpha=0.0)\n",
    "\n",
    "gb.fit(X_train, y_train)\n",
    "y_gb_num = gb.predict(X_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix in array form\n",
      "[[5330    5]\n",
      " [  57    3]]\n",
      "\n",
      "\n",
      "####################\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "####################\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  NO exploit       0.99      1.00      0.99      5335\n",
      "   Exploited       0.38      0.05      0.09        60\n",
      "\n",
      "    accuracy                           0.99      5395\n",
      "   macro avg       0.68      0.52      0.54      5395\n",
      "weighted avg       0.98      0.99      0.98      5395\n",
      "\n"
     ]
    }
   ],
   "source": [
    "utils.metric_evaluation(gb, X_train, X_test, y_test, title = 'GB Tree')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "bst = xgb.XGBClassifier(n_estimators=100,\n",
    "                        colsample_bytree=0.9, \n",
    "                        eta=0.9, \n",
    "                        max_depth=6, \n",
    "                        num_boost_round=10,\n",
    "                        subsample=0.9, \n",
    "                        n_jobs=-1)\n",
    "\n",
    "bst.fit(X_train, y_train)\n",
    "\n",
    "y_xgb_num = bst.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion Matrix in array form\n",
      "[[5329    6]\n",
      " [  53    7]]\n",
      "\n",
      "\n",
      "####################\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "####################\n",
      "\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "  NO exploit       0.99      1.00      0.99      5335\n",
      "   Exploited       0.54      0.12      0.19        60\n",
      "\n",
      "    accuracy                           0.99      5395\n",
      "   macro avg       0.76      0.56      0.59      5395\n",
      "weighted avg       0.99      0.99      0.99      5395\n",
      "\n"
     ]
    }
   ],
   "source": [
    "utils.metric_evaluation(bst, X_train, X_test, y_test, title = 'XGBoost')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Comparing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'ROC Curve')"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(8,8))\n",
    "ax = fig.gca()\n",
    "lr_roc = plot_roc_curve(lr, X_test, y_test, name='LogReg', ax=ax)\n",
    "nb_roc = plot_roc_curve(nb, X_test, y_test, name='NaiveBayes', ax=ax)\n",
    "knn_roc = plot_roc_curve(knn, X_test, y_test, name='KNN', ax=ax)\n",
    "rf_roc = plot_roc_curve(rf, X_test, y_test, name='RandomForest', ax=ax)\n",
    "gb_roc = plot_roc_curve(gb, X_test, y_test, name='GBTree', ax=ax)\n",
    "xgb_roc = plot_roc_curve(bst, X_test, y_test, name='XGBoost', ax=ax)\n",
    "\n",
    "plt.title(\"ROC Curve\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'PRC Curve for Numerical Features')"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(8,8))\n",
    "\n",
    "ax = fig.gca()\n",
    "lr_prc = plot_precision_recall_curve(lr, X_test, y_test, name='LogReg', ax=ax)\n",
    "nb_prc = plot_precision_recall_curve(nb, X_test, y_test, name='NaiveBayes', ax=ax)\n",
    "knn_prc = plot_precision_recall_curve(knn, X_test, y_test, name='NaiveBayes', ax=ax)\n",
    "rf_prc = plot_precision_recall_curve(rf, X_test, y_test, name='RandomForest', ax=ax)\n",
    "gb_prc = plot_precision_recall_curve(gb, X_test, y_test, name='GBTree', ax=ax)\n",
    "xgb_prc = plot_precision_recall_curve(bst, X_test, y_test, name='XGBoost', ax=ax)\n",
    "\n",
    "ax.legend(loc='upper right')\n",
    "plt.title(\"PRC Curve for Numerical Features\")"
   ]
  },
  {
   "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": 2
}