Stock Sentiment Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "import pickle\n",
    "import joblib\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.metrics import confusion_matrix,accuracy_score\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import classification_report\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import plotly.express as px\n",
    "from sklearn.metrics import confusion_matrix\n",
    "import itertools\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>Date</th>\n",
       "      <th>Label</th>\n",
       "      <th>Top1</th>\n",
       "      <th>Top2</th>\n",
       "      <th>Top3</th>\n",
       "      <th>Top4</th>\n",
       "      <th>Top5</th>\n",
       "      <th>Top6</th>\n",
       "      <th>Top7</th>\n",
       "      <th>Top8</th>\n",
       "      <th>...</th>\n",
       "      <th>Top16</th>\n",
       "      <th>Top17</th>\n",
       "      <th>Top18</th>\n",
       "      <th>Top19</th>\n",
       "      <th>Top20</th>\n",
       "      <th>Top21</th>\n",
       "      <th>Top22</th>\n",
       "      <th>Top23</th>\n",
       "      <th>Top24</th>\n",
       "      <th>Top25</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2000-01-03</td>\n",
       "      <td>0</td>\n",
       "      <td>A 'hindrance to operations': extracts from the...</td>\n",
       "      <td>Scorecard</td>\n",
       "      <td>Hughes' instant hit buoys Blues</td>\n",
       "      <td>Jack gets his skates on at ice-cold Alex</td>\n",
       "      <td>Chaos as Maracana builds up for United</td>\n",
       "      <td>Depleted Leicester prevail as Elliott spoils E...</td>\n",
       "      <td>Hungry Spurs sense rich pickings</td>\n",
       "      <td>Gunners so wide of an easy target</td>\n",
       "      <td>...</td>\n",
       "      <td>Flintoff injury piles on woe for England</td>\n",
       "      <td>Hunters threaten Jospin with new battle of the...</td>\n",
       "      <td>Kohl's successor drawn into scandal</td>\n",
       "      <td>The difference between men and women</td>\n",
       "      <td>Sara Denver, nurse turned solicitor</td>\n",
       "      <td>Diana's landmine crusade put Tories in a panic</td>\n",
       "      <td>Yeltsin's resignation caught opposition flat-f...</td>\n",
       "      <td>Russian roulette</td>\n",
       "      <td>Sold out</td>\n",
       "      <td>Recovering a title</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2000-01-04</td>\n",
       "      <td>0</td>\n",
       "      <td>Scorecard</td>\n",
       "      <td>The best lake scene</td>\n",
       "      <td>Leader: German sleaze inquiry</td>\n",
       "      <td>Cheerio, boyo</td>\n",
       "      <td>The main recommendations</td>\n",
       "      <td>Has Cubie killed fees?</td>\n",
       "      <td>Has Cubie killed fees?</td>\n",
       "      <td>Has Cubie killed fees?</td>\n",
       "      <td>...</td>\n",
       "      <td>On the critical list</td>\n",
       "      <td>The timing of their lives</td>\n",
       "      <td>Dear doctor</td>\n",
       "      <td>Irish court halts IRA man's extradition to Nor...</td>\n",
       "      <td>Burundi peace initiative fades after rebels re...</td>\n",
       "      <td>PE points the way forward to the ECB</td>\n",
       "      <td>Campaigners keep up pressure on Nazi war crime...</td>\n",
       "      <td>Jane Ratcliffe</td>\n",
       "      <td>Yet more things you wouldn't know without the ...</td>\n",
       "      <td>Millennium bug fails to bite</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2000-01-05</td>\n",
       "      <td>0</td>\n",
       "      <td>Coventry caught on counter by Flo</td>\n",
       "      <td>United's rivals on the road to Rio</td>\n",
       "      <td>Thatcher issues defence before trial by video</td>\n",
       "      <td>Police help Smith lay down the law at Everton</td>\n",
       "      <td>Tale of Trautmann bears two more retellings</td>\n",
       "      <td>England on the rack</td>\n",
       "      <td>Pakistan retaliate with call for video of Walsh</td>\n",
       "      <td>Cullinan continues his Cape monopoly</td>\n",
       "      <td>...</td>\n",
       "      <td>South Melbourne (Australia)</td>\n",
       "      <td>Necaxa (Mexico)</td>\n",
       "      <td>Real Madrid (Spain)</td>\n",
       "      <td>Raja Casablanca (Morocco)</td>\n",
       "      <td>Corinthians (Brazil)</td>\n",
       "      <td>Tony's pet project</td>\n",
       "      <td>Al Nassr (Saudi Arabia)</td>\n",
       "      <td>Ideal Holmes show</td>\n",
       "      <td>Pinochet leaves hospital after tests</td>\n",
       "      <td>Useful links</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 27 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         Date  Label                                               Top1  \\\n",
       "0  2000-01-03      0  A 'hindrance to operations': extracts from the...   \n",
       "1  2000-01-04      0                                          Scorecard   \n",
       "2  2000-01-05      0                  Coventry caught on counter by Flo   \n",
       "\n",
       "                                 Top2  \\\n",
       "0                           Scorecard   \n",
       "1                 The best lake scene   \n",
       "2  United's rivals on the road to Rio   \n",
       "\n",
       "                                            Top3  \\\n",
       "0                Hughes' instant hit buoys Blues   \n",
       "1                  Leader: German sleaze inquiry   \n",
       "2  Thatcher issues defence before trial by video   \n",
       "\n",
       "                                            Top4  \\\n",
       "0       Jack gets his skates on at ice-cold Alex   \n",
       "1                                  Cheerio, boyo   \n",
       "2  Police help Smith lay down the law at Everton   \n",
       "\n",
       "                                          Top5  \\\n",
       "0       Chaos as Maracana builds up for United   \n",
       "1                     The main recommendations   \n",
       "2  Tale of Trautmann bears two more retellings   \n",
       "\n",
       "                                                Top6  \\\n",
       "0  Depleted Leicester prevail as Elliott spoils E...   \n",
       "1                             Has Cubie killed fees?   \n",
       "2                                England on the rack   \n",
       "\n",
       "                                              Top7  \\\n",
       "0                 Hungry Spurs sense rich pickings   \n",
       "1                           Has Cubie killed fees?   \n",
       "2  Pakistan retaliate with call for video of Walsh   \n",
       "\n",
       "                                   Top8  ...  \\\n",
       "0     Gunners so wide of an easy target  ...   \n",
       "1                Has Cubie killed fees?  ...   \n",
       "2  Cullinan continues his Cape monopoly  ...   \n",
       "\n",
       "                                      Top16  \\\n",
       "0  Flintoff injury piles on woe for England   \n",
       "1                      On the critical list   \n",
       "2               South Melbourne (Australia)   \n",
       "\n",
       "                                               Top17  \\\n",
       "0  Hunters threaten Jospin with new battle of the...   \n",
       "1                          The timing of their lives   \n",
       "2                                    Necaxa (Mexico)   \n",
       "\n",
       "                                 Top18  \\\n",
       "0  Kohl's successor drawn into scandal   \n",
       "1                          Dear doctor   \n",
       "2                  Real Madrid (Spain)   \n",
       "\n",
       "                                               Top19  \\\n",
       "0               The difference between men and women   \n",
       "1  Irish court halts IRA man's extradition to Nor...   \n",
       "2                          Raja Casablanca (Morocco)   \n",
       "\n",
       "                                               Top20  \\\n",
       "0                Sara Denver, nurse turned solicitor   \n",
       "1  Burundi peace initiative fades after rebels re...   \n",
       "2                               Corinthians (Brazil)   \n",
       "\n",
       "                                            Top21  \\\n",
       "0  Diana's landmine crusade put Tories in a panic   \n",
       "1            PE points the way forward to the ECB   \n",
       "2                              Tony's pet project   \n",
       "\n",
       "                                               Top22              Top23  \\\n",
       "0  Yeltsin's resignation caught opposition flat-f...   Russian roulette   \n",
       "1  Campaigners keep up pressure on Nazi war crime...     Jane Ratcliffe   \n",
       "2                            Al Nassr (Saudi Arabia)  Ideal Holmes show   \n",
       "\n",
       "                                               Top24  \\\n",
       "0                                           Sold out   \n",
       "1  Yet more things you wouldn't know without the ...   \n",
       "2               Pinochet leaves hospital after tests   \n",
       "\n",
       "                          Top25  \n",
       "0            Recovering a title  \n",
       "1  Millennium bug fails to bite  \n",
       "2                  Useful links  \n",
       "\n",
       "[3 rows x 27 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('C:/Users/aamish/Desktop/data/Data.csv')\n",
    "df.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# fitting the count vectorizer on the sentences\n",
    "headlines = []\n",
    "\n",
    "cleaned_df = df.copy()\n",
    "\n",
    "cleaned_df.replace('[^a-zA-Z]', ' ',regex=True,inplace=True)\n",
    "cleaned_df.replace('[ ]+', ' ',regex=True,inplace=True)\n",
    "\n",
    "\n",
    "for row in range(len(df)):\n",
    "    headlines.append(' '.join(str(x) for x in cleaned_df.iloc[row,2:]).lower())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CountVectorizer()"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv = CountVectorizer()\n",
    "cv.fit(headlines)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'a hindrance to operations extracts from the leaked reports scorecard hughes instant hit buoys blues jack gets his skates on at ice cold alex chaos as maracana builds up for united depleted leicester prevail as elliott spoils everton s party hungry spurs sense rich pickings gunners so wide of an easy target derby raise a glass to strupar s debut double southgate strikes leeds pay the penalty hammers hand robson a youthful lesson saints party like it s  wear wolves have turned into lambs stump mike catches testy gough s taunt langer escapes to hit  flintoff injury piles on woe for england hunters threaten jospin with new battle of the somme kohl s successor drawn into scandal the difference between men and women sara denver nurse turned solicitor diana s landmine crusade put tories in a panic yeltsin s resignation caught opposition flat footed russian roulette sold out recovering a title'"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "headlines[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4101, 27)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cleaned_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data = cleaned_df[df['Date']<'20150101']\n",
    "test_data = cleaned_df[df['Date']>'20141231']\n",
    "\n",
    "train_data_len = len(train_data)\n",
    "\n",
    "train_headlines = cv.transform(headlines[:train_data_len])\n",
    "test_headlines = cv.transform(headlines[3723:])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3975, 27)\n",
      "(378, 27)\n"
     ]
    }
   ],
   "source": [
    "print(train_data.shape)\n",
    "print(test_data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3975, 46678)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_headlines.shape\n",
    "# train_data['Label'].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(378, 46678)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_headlines.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# lets try with random forest\n",
    "rfc = RandomForestClassifier(n_estimators=200,criterion='entropy')\n",
    "\n",
    "# fitting the model\n",
    "rfc.fit(train_headlines,train_data['Label'])\n",
    "\n",
    "# making predictions\n",
    "preds = rfc.predict(test_headlines)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.8333333333333334\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[140,  46],\n",
       "       [ 17, 175]], dtype=int64)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(accuracy_score(test_data['Label'],preds))\n",
    "confusion_matrix(test_data['Label'],preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "result = []\n",
    "class1 = (test_data['Label'] != preds).sum()\n",
    "class1 = ((len(test_data['Label']) - class1)/len(test_data['Label'])) * 100\n",
    "result.append(class1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Visualization of confusion matrix\n",
    "def plot_confusion_matrix(cm,\n",
    "                          target_names,\n",
    "                          title='Confusion matrix',\n",
    "                          cmap=None,\n",
    "                          normalize=True):\n",
    "    accuracy = np.trace(cm) / float(np.sum(cm))\n",
    "    misclass = 1 - accuracy\n",
    "\n",
    "    if cmap is None:\n",
    "        cmap = plt.get_cmap('Blues')\n",
    "\n",
    "    plt.figure(figsize=(8, 6))\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "\n",
    "    if target_names is not None:\n",
    "        tick_marks = np.arange(len(target_names))\n",
    "        plt.xticks(tick_marks, target_names, rotation=45)\n",
    "        plt.yticks(tick_marks, target_names)\n",
    "\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "\n",
    "\n",
    "    thresh = cm.max() / 1.5 if normalize else cm.max() / 2\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        if normalize:\n",
    "            plt.text(j, i, \"{:0.4f}\".format(cm[i, j]),\n",
    "                     horizontalalignment=\"center\",\n",
    "                     color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "        else:\n",
    "            plt.text(j, i, \"{:,}\".format(cm[i, j]),\n",
    "                     horizontalalignment=\"center\",\n",
    "                     color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label\\naccuracy={:0.4f}; misclass={:0.4f}'.format(accuracy, misclass))\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "cm =  confusion_matrix(test_data['Label'],preds)\n",
    "\n",
    "plot_confusion_matrix(cm = cm, \n",
    "                      normalize    = False,\n",
    "                      target_names = ['Stock_UP','Stock_DOWN'],\n",
    "                      title        = \"Confusion Matrix of Random Forest\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "********************Training**********************\n"
     ]
    }
   ],
   "source": [
    "from sklearn import svm\n",
    "clf = svm.SVC(kernel='rbf', random_state=1, gamma='auto', C=1.0)\n",
    "#\n",
    "# training\n",
    "print(\"********************Training**********************\")\n",
    "clf.fit(train_headlines, train_data['Label'])\n",
    "\n",
    "preds = clf.predict(test_headlines)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5079365079365079\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[  0, 186],\n",
       "       [  0, 192]], dtype=int64)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(accuracy_score(test_data['Label'],preds))\n",
    "confusion_matrix(test_data['Label'],preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "class1 = (test_data['Label'] != preds).sum()\n",
    "class1 = ((len(test_data['Label']) - class1)/len(test_data['Label'])) * 100\n",
    "result.append(class1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfYAAAHCCAYAAAAdAOsHAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAA4OUlEQVR4nO3dZ7hkVZm38ftPE0RAyQxBJIOAitIwioIgOoAyiAnBMBiRV9Axi2EwYsKcs4AiICJKUIIIIihKk4OiKDo0MEhQEIndPO+HvU+zOZzUp/t0naq+f151ddXaae1q7KeetdZeK1WFJEkaDEv0ugKSJGnhMbBLkjRADOySJA0QA7skSQPEwC5J0gAxsEuSNEAM7NJCkmTZJCcmuS3JsQtwnpckOW1h1q0Xkvw0yb5TcN7nJrk2yR1JnrCwzy/1OwO7FjtJXpxkVhsYbmgD0FMXwqlfAKwBrFJVL5zsSarqyKr6j4VQnwdJsmOSSvLDYeWPb8vPmuB53pfku+PtV1W7VdXhk6zuWD4BHFhVy1fVRSPU7zlJLk5ye5Kbk5yRZL0k+yT5S5IM23/JJH9LsvvC+o6kXjKwa7GS5M3AZ4AP0wThdYEvAc9ZCKd/NPCHqpqzEM41VW4CtkuySqdsX+APC+sCaUzlvy2PBq4Y5dobAUcAbwEeCaxP8/d7P3A8sCLwtGGH7QoUcEr7ecq/I2kqGdi12EjySOADwAFV9cOq+ldV3VdVJ1bV29p9lknymSTXt6/PJFmm3bZjktlJ3tJmeDckeUW77f3AwcCL2paAVw3PbNussZIs2X5+eZI/J/lnkmuSvKRTfk7nuO2SnN828Z+fZLvOtrOSfDDJue15Tkuy6hhfw73Aj4C92+NnAHsBRw77rj7bNnffnuSCJNu35bsC7+rc5yWdehyS5FzgTmCDtuzV7fYvJ/lB5/wfazPpB2XP7bYlkrwnyV/b7/mIJI9s/27uAGYAlyT50wj3txVwTVWdUY1/VtVxVfW/VXU38H3gv4Yd81/AkZ0fZBP6jqTpysCuxcmTgYfRZG6jeTfwJJoA8XhgW+A9ne3/RpMJrg28CvhikpWq6r00rQDHtE3E3xyrIkmWAz4H7FZVKwDbARePsN/KwMntvqsAnwJOHpZNvhh4BbA6sDTw1rGuTZPRDgW3XWiy3+uH7XM+zXewMvA94NgkD6uqU4bd5+M7x7wM2A9YAfjrsPO9BXhc+6Nle5rvbt8aeU7rl7evnYANgOWBL1TVPVW1fLvP46tqwxGOvRDYLMmnk+yUZPlh2w8HXpBkWZj3Y+8/2++kayLfkTQtGdi1OFkFuHmcpvKXAB+oqr9V1U3A+2kC1pD72u33VdVPgDuATSdZn/uBLZMsW1U3VNVIzcvPBv5YVd+pqjlVdRTwe5pgNOTbVfWHqrqLJiPdaqyLVtWvgJWTbEoTvIYHNarqu1V1S3vNTwLLMP59HlZVV7TH3DfsfHcCL6X5YfJd4PVVNXuU87wE+FRV/bmq7gDeCew91NIxzr39GdiR5ofX94Gbkxw2FOCr6lzgRuC57SF70XSfXDzsPON+R9J0ZWDX4uQWYNVxAsRaPDjb/GtbNu8cw34Y3EmTUc6XqvoX8CJgf+CGJCcn2WwC9Rmq09qdz/83ifp8BziQJit+SAtG293wu7b5/x80rRRjNfEDXDvWxqr6LfBnIDRBdzQj/R0sSTMmYlxVdV5V7VVVqwHbAzvQtMQM6WbjL6PJ4kcy5nckTVcGdi1Ofg3cDew5xj7X0wzOGrIuk2+C/Rfw8M7nf+turKpTq+qZwJo0WfjXJ1CfoTpdN8k6DfkO8DrgJ202PU/bVP4Ommx2papaEbiNJiBDM9BsJGMuFZnkAJrM/3rg7WPsOtLfwRyaTHu+VNX5wA+BLTvFRwA7J3kyTbfL90Y5fNTvSJrODOxabFTVbTQD3L6YZM8kD0+yVJLdkny83e0o4D1JVmsHoR1M03Q8GRcDOyRZt+3LfefQhiRrJNmj7Wu/h6ZJf+4I5/gJsEmaR/SWTPIiYHPgpEnWCYCquoZmdPi7R9i8Ak0gvQlYMsnBwCM6228E1pufke9JNgE+RNMc/zLg7Um2GmX3o4A3JVm/bUIf6tMf92mDJE9N8pokq7efNwP2AM4b2qeq/gqc017n9Kr6v5HONc53JE1bBnYtVqrqU8CbaQbE3UTTfHwgzShoaILPLOBS4DKawVgfmuS1TgeOac91AQ8OxkvQDCi7HriVJoC8boRz3ALs3u57C02mu3tV3TyZOg079zlVNVJrxKnAT2ke7/orTStHt5l9aPKdW5JcON512q6P7wIfq6pLquqPNCPrvzP0xMEw36LJls8Grmmv//qJ3RX/oAnkl7Uj6E+haUb/+LD9DqdpFRiz73yM70iatjLyoFRJktSPzNglSRogBnZJkgaIgV2SpAFiYJckaYAY2CVJGiDjTtGohW/VVVetRz96vV5XQ3qIi64abZZXqbfqXzfe3M4mOKVmPOLRVXPumvTxdddNp1bVrguxSvPNwN4Dj370epz7m1m9rob0ECs9dawJ4aTeufs3hw6fWnlK1Jy7WWazvSd9/N0XfX68qZennIFdkqQhAR66mnBfsY9dkqQBYsYuSVLXxJdBmJYM7JIkdfV5U7yBXZKkedL3GXt/116SJD2IGbskSV02xUuSNCBC3zfFG9glSZonfZ+x9/fPEkmS9CBm7JIkddkUL0nSAOnzpngDuyRJ8/T/c+wGdkmShrgIjCRJmk7M2CVJ6rIpXpKkQWEfuyRJg2UJ+9glSdI0YcYuSdIQ54qXJGnA9PnjbgZ2SZLm6f/Bc/1de0mS9CBm7JIkddkUL0nSAJnCpvgk3wJ2B/5WVVu2ZccAm7a7rAj8o6q2SrIe8DvgqnbbeVW1/3jXMLBLkjQkmeqM/TDgC8ARQwVV9aIHLp9PArd19v9TVW01PxcwsEuS1DWFGXtVnd1m4g+9bBJgL+DpC3INB89JkrTwrJpkVue133wcuz1wY1X9sVO2fpKLkvwiyfYTOYkZuyRJXQvWFH9zVc2c5LH7AEd1Pt8ArFtVtyTZGvhRki2q6vaxTmJglyRpnt48x55kSeB5wNZDZVV1D3BP+/6CJH8CNgFmjXUuA7skSV29edztGcDvq2r2A9XIasCtVTU3yQbAxsCfxzuRfeySJC0iSY4Cfg1smmR2kle1m/bmwc3wADsAlya5BPgBsH9V3TreNczYJUkaMsWLwFTVPqOUv3yEsuOA4+b3GgZ2SZLm6f+54g3skiR19fmUsv39s0SSJD2IGbskSV02xUuSNED6vCnewC5J0pA4eE6SpMHS5xl7f/8skSRJD2LGLklSR/o8YzewS5LUCgZ2SZIGR9pXH7OPXZKkAWLGLknSPLEpXpKkQWJglyRpgPR7YLePXZKkAWLGLklSR79n7AZ2SZKGDMDjbgZ2SZJaGYBR8faxS5I0QMzYJUnq6PeM3cAuSVKHgV2SpAFiYJckaVAMwKh4B89JkjRAzNglSeqwKV6SpAExCM+xG9glSero98BuH7skSQPEjF2SpK7+TtgN7JIkzZP+b4o3sEuS1NHvgd0+dkmSBogZuyRJHWbskiQNiKHn2Cf7Gvf8ybeS/C3J5Z2y9yW5LsnF7etZnW3vTHJ1kquS7DKRezCwS5LUlQV4je8wYNcRyj9dVVu1r58AJNkc2BvYoj3mS0lmjHcBA7skSUPaUfFTlbFX1dnArROszXOAo6vqnqq6Brga2Ha8gwzskiT13oFJLm2b6ldqy9YGru3sM7stG5OBXZKkjgXM2FdNMqvz2m8Cl/wysCGwFXAD8Mmhqoywb413MkfFS5LUsYCj4m+uqpnzc0BV3di59teBk9qPs4FHdXZdB7h+vPOZsUuS1DW1g+ceerlkzc7H5wJDI+ZPAPZOskyS9YGNgd+Odz4zdvWt0049hbe++b+ZO3cuL3/lq3nb2w/qdZW0GPnKu1/Ibk95DDf9/Q5mvuRTADxu4zX5/DuexzJLL8WcuffzxkOPZ9aVTRfplhv9G194x/NZYblluP/+4qmv/Dz33Dunl7egHkhyFLAjTZP9bOC9wI5JtqJpZv8L8FqAqroiyfeBK4E5wAFVNXe8axjY1Zfmzp3LG99wACf/9HTWXmcdnvqkbdh99z14zOab97pqWkx85+RZfOUHv+IbB79oXtkhBz6bQ775M0779VXs8uTNOOTAZ7HL677KjBlL8K337cOr3nc0l119Ays/4uHcN2fcf5/VI1M5QU1V7TNC8TfH2P8Q4JD5uYZN8epL5//2t2y44Uasv8EGLL300rzwRXtz0ok/7nW1tBg59+JruPX2Ox9UVlU8YrmHAfDI5R/GDTfdDsAztt2Ey6++gcuuvgGAW2+/k/vvH3cMlHpgQQbOTZcZ68zY1Zeuv/461lnngTEla6+9Dr/97W96WCMJ3vaZEznxM6/iI69/Nksk7LTfFwHYeN1VqSpO+MyrWHWl5fnB6Rfzqe/+ose11WimS4CerL7J2JO8O8kV7XN+Fyf59yRvTPLwSZ7vfUneOsF9z0oys/N5vaHpAJPsmOS2JBcl+V2S906mPpo/VQ/Ndvr9/4zqf/s970m8/bMnsvFzPszbP3siX373CwFYcsYSbPf49XnFe49i5/2+xB5P25IdZ27U49pqUPVFYE/yZGB34IlV9TjgGTQP7b8RmFRgX8h+WVVPAGYCL02yda8rNOjWXnsdZs9+YN6G666bzVprrdXDGknwkmdtzY/ObAY0H3fGpczcvGlVuu5vt/HLi/7MLbfdyV333Mcpv/o9T9h03HlG1CP93hTfF4EdWJPm2cB7AKrqZuAFwFrAmUnOBEiyT5LLklye5GNDByfZNcmFSS5Jcsbwkyd5TZKfJll2QSpZVf8CLqCZaEBTaOY223D11X/kL9dcw7333suxxxzNs3ffo9fV0mLuhptvZ/snbgDAjjM34uprbwbg9N/8gS03WpNll1mKGTOWYPsnbsDvrrlxrFOplxbx424LW7/0sZ8GHJzkD8DPgGOq6nNJ3gzsVFU3J1kL+BiwNfB34LQkewLnAl8Hdqiqa5Ks3D1xkgOB/wD2HPrhMFlJVgGeBHxwhG37AfsBPGrddRfkMgKWXHJJPv3ZL/Cfz96FuXPnsu/LX8nmW2zR62ppMXL4B17M9k/cgFVXXI6rT3gXH/z66RzwkeM49E17sOSMJbjn3jkc+JHjAPjHP+/ic0edzTnffj1VcOqvf88pv/p9j+9Ao5kumfdk9UVgr6o72ubt7YGdgGOSDH9oeRvgrKq6CSDJkcAOwFzg7HYCfaqqO/n+y2hm9tmzqu4bqwrjlG2f5CLgfuCjVXXFCPfwNeBrAFtvPdPhsAvBrrs9i113e9b4O0pTYN+Dvzdi+VNe/rkRy48+5SKOPuWiqaySFoYY2BeZ9qH8s4CzklwG7Dtsl9H+JsLoc+teTjM37zrANWNc/hZgpc7nlYGbO59/WVW7j3G8JEmLRF/0sSfZNMnGnaKtgL8C/wRWaMt+Azwtyapp1qvdB/gF8Ou2fP32XN2m+ItoZvg5oW3KH81ZNIPihn487AucuUA3JUmadgIkk39NB/2SsS8PfD7JijTT6l1N01+9D/DTJDdU1U5J3kkTcAP8pKp+DPP6t3+YZAngb8Azh05cVee0j72dnOSZ7cC84b4GbAZckqSAWcA7p+heJUk9M31Gt09WXwT2qroA2G6ETZ9vX0P7fQ94SMdXVf0U+Omwsvd13p8KnDrG9e8FDhxl21k0Gb0kaQD0eVzvj6Z4SZI0MX2RsS8qSb4IPGVY8Wer6tu9qI8kadGzKX6AVNUBva6DJKmHptEguMkysEuS1AqwxBL9HdntY5ckaYCYsUuS1GFTvCRJA8TBc5IkDYoBGDxnH7skSQPEjF2SpFYzV3x/p+wGdkmS5nGueEmSBkqfx3UDuyRJXf2esTt4TpKkAWLGLknSkAF43M3ALklSy1HxkiQNmD6P6/axS5I0SMzYJUnqsClekqQB0udx3cAuSdI86f+M3T52SZIGiBm7JEmt5nG3XtdiwRjYJUmap/8XgbEpXpKkjmTyr/HPnW8l+VuSyztlhyb5fZJLkxyfZMW2fL0kdyW5uH19ZSL1N7BLktSRZNKvCTgM2HVY2enAllX1OOAPwDs72/5UVVu1r/0ncgEDuyRJi0hVnQ3cOqzstKqa0348D1hnQa5hYJckacgCNMO3CfuqSWZ1XvvNZw1eCfy083n9JBcl+UWS7SdyAgfPSZLUWgiLwNxcVTMnde3k3cAc4Mi26AZg3aq6JcnWwI+SbFFVt491HgO7JEkdvRgVn2RfYHdg56oqgKq6B7infX9Bkj8BmwCzxjqXTfGSJPVQkl2BdwB7VNWdnfLVksxo328AbAz8ebzzmbFLktQxlQl7kqOAHWn64mcD76UZBb8McHrbWnBeOwJ+B+ADSeYAc4H9q+rWEU/cYWCXJKljKpviq2qfEYq/Ocq+xwHHze81DOySJA2Z4EQz05l97JIkDRAzdkmSWhmAueIN7JIkdfR5XDewS5LUtUSfR3YDuyRJHX0e1x08J0nSIDFjlySp1Szm0t8pu4FdkqSOJfo7rhvYJUnq6veM3T52SZIGiBm7JEkdfZ6wG9glSRoSmtnn+pmBXZKkjn4fPGcfuyRJA8SMXZKkIXERGEmSBkqfx3UDuyRJQ0L/LwJjH7skSQPEjF2SpI4+T9gN7JIkdTl4TpKkAdGs7tbrWiyYUQN7ks8DNdr2qnrDlNRIkqQe6vfBc2Nl7LMWWS0kSdJCMWpgr6rDu5+TLFdV/5r6KkmS1Dv9na9P4HG3JE9OciXwu/bz45N8acprJklSD6SdfW4yr+lgIs+xfwbYBbgFoKouAXaYwjpJktQTzQQ1k39NBxOaoKaqrh1WNHcK6iJJkhbQRB53uzbJdkAlWRp4A22zvCRJA2UaNalP1kQC+/7AZ4G1geuAU4EDprJSkiT1Sp/H9fEDe1XdDLxkEdRFkqSe6/eMfSKj4jdIcmKSm5L8LcmPk2ywKConSZLmz0QGz30P+D6wJrAWcCxw1FRWSpKkXlhcRsWnqr5TVXPa13cZY6pZSZL62cA+x55k5SQrA2cmOSjJekkeneTtwMmLroqSJC06WYDXuOdOvtV2a1/eKVs5yelJ/tj+uVJn2zuTXJ3kqiS7TKT+Yw2eu4AmMx+q62s72wr44EQuIElSv0imfBGYw4AvAEd0yg4CzqiqjyY5qP38jiSbA3sDW9B0hf8sySZVNeZcMmPNFb/+AlZekiR1VNXZSdYbVvwcYMf2/eHAWcA72vKjq+oe4JokVwPbAr8e6xoTWo89yZbA5sDDOpU7YvQjJEnqTz3oKl+jqm4AqKobkqzelq8NnNfZb3ZbNqZxA3uS99L8ktgc+AmwG3AOD25GkCRpICzgILhVk3SXPf9aVX1tslUZoWzcwesTydhfADweuKiqXpFkDeAb81k5SZL6wgJm7DdX1cz5PObGJGu22fqawN/a8tnAozr7rQNcP97JJvK4211VdT8wJ8kj2gs6QY0kSQvHCcC+7ft9gR93yvdOskyS9YGNgd+Od7KJZOyzkqwIfJ1mpPwdEzmxJEn9JmRKR8UnOYqme3vVJLOB9wIfBb6f5FXA/wIvBKiqK5J8H7gSmAMcMN6IeJjYXPGva99+JckpwCOq6tJJ3I8kSdNbpnbwXFXtM8qmnUfZ/xDgkPm5xqiBPckTx9pWVRfOz4UkSeoH02UGuckaK2P/5BjbCnj6Qq6LpF67585e10DSAhprgpqdFmVFJEmaDiYyqnw6m9AENZIkLQ7CYDfFS5K02Jkuy69OloFdkqSOfg/s43YlpPHSJAe3n9dNsu3UV02SJM2viYwR+BLwZGDo2bt/Al+cshpJktQjSdPHPtnXdDCRpvh/r6onJrkIoKr+nmTpKa6XJEk90e9N8RMJ7PclmUG7okyS1YD7p7RWkiT1yDRJvCdtIk3xnwOOB1ZPcgjNkq0fntJaSZKkSZnIXPFHJrmAZh7bAHtW1e+mvGaSJC1igSldBGZRGDewJ1kXuBM4sVtWVf87lRWTJKkXFoeZ506m6V8P8DBgfeAqYIsprJckST3R5wn7hJriH9v93K769topq5EkSZq0+Z55rqouTLLNVFRGkqReSrJY9LG/ufNxCeCJwE1TViNJknqoz+P6hDL2FTrv59D0uR83NdWRJKm3BnqCmnZimuWr6m2LqD6SJGkBjBrYkyxZVXPawXKSJA28QX+O/bc0/ekXJzkBOBb419DGqvrhFNdNkqRFrs/j+oT62FcGbgGezgPPsxdgYJckDZYMdh/76u2I+Mt5IKAPqSmtlSRJPRL6O7KPFdhnAMvDiHdoYJckaRoaK7DfUFUfWGQ1kSSpx5rBc72uxYIZK7D3+a1JkjT/Bjmw77zIaiFJ0jSRPh8WP+rqdFV166KsiCRJWnDzvQiMJEmDatD72CVJWrxk8ZigRpKkxUa/Tyk7ah+7JEnqP2bskiS17GOXJGnA9HlLvIFdkqQHhCWmcH62JJsCx3SKNgAOBlYEXgPc1Ja/q6p+MplrGNglSWqFqc3Yq+oqYCuAJDOA64DjgVcAn66qTyzoNRw8J0lSb+wM/Kmq/rowT2pglyRpSLse+2Rf82lv4KjO5wOTXJrkW0lWmuwtGNglSepYIpn0C1g1yazOa7+RrpFkaWAP4Ni26MvAhjTN9DcAn5xs/e1jlySptRD62G+uqpkT2G834MKquhFg6E+AJF8HTppsBczYJUla9Pah0wyfZM3OtucCl0/2xGbskiR1TPWUskkeDjwTeG2n+ONJtgIK+MuwbfPFwC5JUsdUT1BTVXcCqwwre9nCOr+BXZKkVuj/Pup+r78kSeowY5ckaUggfT5ZvIFdkqSO/g7rBnZJkuZplm3t79BuYJckqaO/w7qD5yRJGihm7JIkdfR5S7yBXZKkB8RR8ZIkDQonqJEkSdOKGbskSR02xUuSNED6O6wb2CVJesAATClrH7skSQPEjF2SpNYgjIo3sEuS1NHvTfEGdkmSOvo7rPd/i4MkSeowY5ckqaPPW+IN7JIkDWkGz/V3ZDewS5LUYcYuSdLACOnzjN3Bc5IkDRAzdkmSOmyKlyRpQDh4TpKkQZL+z9jtY5ckaYCYsUuS1NHvGbuBXZKkjn5/3M3ALklSK8AS/R3X7WOXJGmQmLFLktRhU7wkSQOk3wfP2RSvvnXaqafwuC02ZYvNNuLQj3+019XRYuYr730Jfz3jI8w69l3zyh67ydqcdfhbOP/77+IHn3ktKyz3MACe/u+bce6Rb+f877+Lc498O0/bZpNeVVsTkAX433RgYFdfmjt3Lm98wwH8+MSfctGlV3Ls0Ufxuyuv7HW1tBj5zonn8ZwDvvigsi8f/GLe87kfs81eH+aEMy/hTfvuDMAt/7iDF7zxq2yz14d5zcHf4Vsf+q9eVFkTMDR4brKvCV0j+UuSy5JcnGRWW7ZyktOT/LH9c6XJ3oOBXX3p/N/+lg033Ij1N9iApZdemhe+aG9OOvHHva6WFiPnXvgnbr3tzgeVbfzo1TnngqsB+Pl5v2fPnbcC4JKrZnPDTbcBcOWfbmCZpZdi6aXsCV3M7VRVW1XVzPbzQcAZVbUxcEb7eVIM7OpL119/Heus86h5n9deex2uu+66HtZIaoL27js+FoDnPfOJrLPGQ5Ou5z5jKy656lruvW/Ooq6eJmRBGuIXqCn+OcDh7fvDgT0neyIDu/pSVT2kLP0+4kV977XvO5LX7rUD5x75dpZ/+DLce9/cB21/zAb/xofe8BwO/NDRPaqhxtXOFT/ZF7Bqklmd134jXKWA05Jc0Nm+RlXdAND+ufpkb2FK24KSvBt4MTAXuB94LfBk4GtVdedYx45yvvcBd1TVJyaw72HA04DbgWWB84B3VtV17fZHAp8HntIeci7w+qq6LcnxwOFV9aN236uA71TVh9rPxwFHArcCZwJ7VNWJ7baTgE9U1Vnze3+auLXXXofZs6+d9/m662az1lpr9bBGEvzhLzfyn69r+t03Wnd1dtt+i3nb1l59RY751H68+n++wzWzb+5VFTUBC5gi3NxpXh/NU6rq+iSrA6cn+f2CXfLBpixjT/JkYHfgiVX1OOAZwLXAG4GHT9V1h3lbVT0e2BS4CDgzydLttm8Cf66qDatqQ+Aa4Bvttl8B27X3sQpwB80PkiFPbvcBmA28e0rvQg8xc5ttuPrqP/KXa67h3nvv5dhjjubZu+/R62ppMbfaSssDTevRQa/Zha//4BwAHrn8svzw8/tz8OdP4NeX/LmXVdQ0UFXXt3/+DTge2Ba4McmaAO2ff5vs+aeyKX5Nml8u9wBU1c3AC4C1aALsmQBJ9mlHB16e5GNDByfZNcmFSS5Jcsbwkyd5TZKfJll2vIpU49PA/wG7JdkI2Br4YGe3DwAzk2xIk71v15ZvB5wErJbG+sBdVfV/7fZLgNuSPHPiX40W1JJLLsmnP/sF/vPZu7DVYx/D81+4F5tvscX4B0oLyeEfeTlnHf4WNnn0Glx9ygfZd88ns9euM7n0RwdzyfH/ww033cYRPz4PgP333oENH7UaB71mV847+iDOO/qgeT8CNL00o+Iz6de450+WS7LC0HvgP4DLgROAfdvd9gUmPRp4KpviTwMOTvIH4GfAMVX1uSRvphkNeHOStYCP0QTZv9P0OexJE1i/DuxQVdckWbl74iQH0nwZew79cJigC4HNaPo3Lq6qeR1gVTU3ycXAFsCpwJZtdr8d8AtgA+AxwBPa+nV9qH2dPtqF236U/QAete6681FljWbX3Z7Frrs9q9fV0GJq33ceNmL5F4866yFlH/vGqXzsG6dObYW00EzxaJ01gOPbMUFLAt+rqlOSnA98P8mrgP8FXjjZC0xZYK+qO5JsDWwP7AQck2T48P1tgLOq6iaAJEcCO9D0yZ9dVde057q1c8zLaJq/96yq++azWun8+dDRV215Vd2T5ArgicCTgI/TBPbtaAL7r7oHVdUvk5Bk+9EuXFVfA74GsPXWM0e6tiRpOpjCyF5VfwYeP0L5LcDOC+MaUzoqvqrmVtVZVfVe4EDg+cN2Ge3rGy3wQtNksR6wziSq9ATgd8AVwBOSzLv/9v3j2+3QBO8dgBWq6u80g++2a1/DM3aAQ7CvXZLUY1M5eG7TJBt3irYC/gr8E1ihLfsN8LQkqyaZAexD0+z967Z8/fZc3ab4i2hG15/QNuVPpC5J8gaafv9Tqurq9jzv6ez2HuDCdhs0wfu1NH3oAJfSZO/r0vwweJCqOg1YiRF+iUmS+odTyo5ueeDwJFcmuRTYHHgfTXP0T5Oc2T6r906aR8YuoQmsP26b5vcDfpjkEuCY7omr6hzgrcDJSVYdow6Htsf/gabZf6equrfd9ipgkyRXJ/kTsElbNuRXNM3vv26vOYdmlOKsqrp/lOsdwuRaEiRJ08QCPsfecxlpog9Nra23nlnn/mZWr6shPcRK2xzY6ypII7r74i9eMIHnwxfYYx77hDrix2dN+vhtN1xxkdRzLE5WLElS1zTJvCer7wN7ki/ywOxxQz5bVd/uRX0kSeqlvg/sVXVAr+sgSRoMgWkzCG6y+j6wS5K00EyjQXCTZWCXJKmjz+O6y7ZKkjRIzNglSerq85TdwC5J0jzTZwa5yTKwS5LU0e+D5+xjlyRpgJixS5LUCn3fxW5glyTpQfo8shvYJUnq6PfBc/axS5I0QMzYJUnq6PdR8QZ2SZI6+jyuG9glSZpnAIbFG9glSepw8JwkSZo2zNglSWoFB89JkjRQ+jyuG9glSXqQPo/s9rFLkjRAzNglSero91HxBnZJkjocPCdJ0gDp87huH7skSYPEjF2SpK4+T9kN7JIktZqp4vs7shvYJUkakv4fPGcfuyRJHVmA17jnTh6V5Mwkv0tyRZL/bsvfl+S6JBe3r2dNtv5m7JIkLTpzgLdU1YVJVgAuSHJ6u+3TVfWJBb2AgV2SpK4pbIqvqhuAG9r3/0zyO2DthXkNm+IlSZonC/S/+bpSsh7wBOA3bdGBSS5N8q0kK032DgzskiR1JJN/AasmmdV57TfyNbI8cBzwxqq6HfgysCGwFU1G/8nJ1t+meEmSFp6bq2rmWDskWYomqB9ZVT8EqKobO9u/Dpw02QqYsUuS1FqQEfETHBUf4JvA76rqU53yNTu7PRe4fLL3YMYuSVLX1D7H/hTgZcBlSS5uy94F7JNkK6CAvwCvnewFDOySJHVM5cxzVXUOI/90+MnCuoZN8ZIkDRAzdkmSOvp9SlkDuyRJHX0e1w3skiTNMwCLwBjYJUl6kP6O7A6ekyRpgJixS5LUCjbFS5I0UPo8rhvYJUnq6veM3T52SZIGiBm7JEkdUzml7KJgYJckqau/47qBXZKkrj6P6/axS5I0SMzYJUlqxSllJUkaLA6ekyRpkPR3XLePXZKkQWLGLklSR58n7AZ2SZK6HDwnSdLAiIPnJEkaFIOwbKuD5yRJGiAGdkmSBohN8ZIkdfR7U7yBXZKkjn4fPGdTvCRJA8SMXZKkIS4CI0nS4AjOPCdJ0mDp88huH7skSQPEjF2SpI5+HxVvYJckqcPBc5IkDZA+j+sGdkmSHqTPI7uD5yRJWoSS7JrkqiRXJzloYZ/fjF2SpI6pHDyXZAbwReCZwGzg/CQnVNWVC+saZuySJLWG1mOf7GsCtgWurqo/V9W9wNHAcxbmPZix98CFF15w87JL5a+9rscAWRW4udeVkEbgf5sLz6MXxUUuvPCCU5ddKqsuwCkelmRW5/PXquprnc9rA9d2Ps8G/n0BrvcQBvYeqKrVel2HQZJkVlXN7HU9pOH8b7P/VNWuU3yJkfL6WpgXsClekqRFZzbwqM7ndYDrF+YFDOySJC065wMbJ1k/ydLA3sAJC/MCNsVrEHxt/F2knvC/TT1IVc1JciBwKjAD+FZVXbEwr5Gqhdq0L0mSesimeEmSBoiBXZKkAWJglyRpgBjYJUkaIAZ2LTaSrJlkxySrtPM1S9NekjWSPCaJ/15rQnzcTYuFJJsB3wduaos+y0J+dlSaIm8C1gU+nOSK8lEmjcNfgBp4STakeWb0nVW1M3Aa8OYky/e2ZtKEvIvmB+l/A1smE1xqRIstA7sWB48B7gCWBqiqjwF3s4gWlZDmV5KNkzwvyeOr6n7gjTSLybwReKxdSRqLTfEaWEnWA9atqpPa/snXtH8uAywP3NDL+kkjSbIUTZa+L3Btkq8A/wI+DHwK2As4BrisZ5XUtGZg1yDbGvhckpdU1QltUD+Ipr9ys6q6PcmSVTWnt9WUGknWAB4HHALcSLOc5w3tn18ClgKeCTwxyVuq6ne9qqumLwO7BlZVHddmP59P8t9V9aM2uL8W2BE4waCuaeZlwG5VtXOSL9Is8flkYH+aVcCeCCwHPAVYuWe11LTmXPEaKEkeAdxXVXd1yl4MvBP476r6eZI9aAYifbeqvt2jqkrzJFmqqu5r358OXFdVL0/yaOBtwMOAt1bVP5IsB8yoqtt7WGVNYw6e08BIsizwHeAVSR4+VF5V36Ppn3x/ki2Ak4DPYx+lpoEk6wOHJPmPtuilwHJJ9qqqvwKfAf4BfDXJI6vqXwZ1jcXAroHRZulfBJ4L7N0N7sCxwK+Ame0o4xOralYPqikNNxN4HXBoklcCdwLnAFu0LVB/Ar4J/B8+yaEJsI9dA6WqTktyL/BegCRHV9Wd7RrI/wLWbPeb28t6SkMDN6vq2La16XHA7jSPZS4D7AJcUVXfT/J7mnkY7uxhldUnDOwaOFV1VjuHx/8AM5JcBNwLPB94fS/rJgEkWQt4bZKLqupHwG+AlYAjaLLyGTQ/Qr+V5Nqq+jVNJi+Ny8CugdQG93uBvYGnA6sD/1NVZ/e2ZtI8ywP/meRZVbVfkucDL6yq/4F5Y0b+H82PUmnCHBWvvpUk482bPTRKHlipqq6fyDHSVGmfU18emFtVf0myMfARmmz9/TSDPI+uqi+0+69TVbN7VmH1JQO7+k43OCd5Gs2jPz8fb9/28xLt4DlpkUqyJfBtmsFwGwOXAh+tqquSvBtYBXgUsBbwmqq6smeVVV9zVLz6SpLVgZcmWbEt2gMYK1Av0R43A8Cgrl5Isi7NNLCHVtXewPbAXOCtSbaoqkOA44CLaCakcS54TZqBXf1mR+BpwPOSLEPz3/AjoMnGuytfJZlRVXPbHwEf7PwYkBa1LYDz2hHuy7Sj218PzAHeA1BV51bVh4G1qso5FjRpBnb1haGAXVXfB34GbEMzMG4Jmj70oWx8xXb/oaD+SODHwKlV9Y9FX3Mtzjo/NJej/W+zqu5pH3W7i+b59U2TPL1z2I3DjpXmi4FdfSXJRlV1NHAusB2wK/DNJIclORv4eZL1O5n6j4B3V9UvelZpLbY64zvOATZuJ6ChnVdh+Xb7RcCtw49xkKcmy8fd1BeqqpLsCnwmyQuq6rtJ7gEKuB74Os2a6yu0o9+XAb4KvLeqzuldzbU4SrIhzWQzmwKzgFNpmtyfn2ROVR1RVXe0UxxvS9PfLi0UjopXX0iyEXACzWjhczvlL6UZiDQL+MawEfA+KqRFLslmwA+A42lWZ9sOuJBmbYI7aCZOugKYDewJvKuqjutJZTWQDOya1oYeV0uyFc0/gHu15ctU1T3t+xcAv6+qy7vH9KzSWmy1z6UfB3ykqo5qy9YDXkOzQttXgLtpAvrfgWuq6lz/m9XCZGDXtNQJ6Eu2/ZErAScDX2hXayPJbsDWVfWhnlZWoll6FfgczSyHL6+qfw7Nm9Cu4PYF4OSq+lJPK6qBZ2DXtNMJ6k+nyWxmA+cDK9NMD/t34BfAocB7quqkXtVV6kryROAA4PfAj6rqj51tbwZ2qqr/dKIkTSVHxWvaaYP6DsC3aEYMr0szEGkT4DBgM+DZNHO/n+RjQZouqupCmqWDtwD2bMeGDPkncEG7n0FdU8aMXdNSkgOAJarq80lWoZmUZmfgTTzw3HrZN6npqM3c3wBcCXyZZnT8kcABVfWzXtZNg8+MXT2XZLkk27Tvn94+AnQb8LIka1XVLcDpwGOAjasFPuur6anN3D9HE9APpQnqbzaoa1EwsKunkqxJ08T+3iTfAz4DrEAzsvhUmrm0HwWsQTN71309qqr0ION1AXWa5R8BvL6qTl4kFdNiz6Z49Uz7vO/xwCtpVrX6KnB8Vb2y/Udza2Avmib4u4BP+7yvem1+Vhds91m+qu5o3ztoTlPOwK6eaJ/tPYkmWH+zfRxoNZoZ5I6qqo929l0aeGRV3WSfunqpXV1wF+DEqvpHkk+2788aZf+hNQuWrKo5i7KuWnzZFK9e2Qk4ow3qAR5Js0jGMcB/J3l1kl2SnAksW1U3gX3q6rkdmdzqgh9wdUEtKs4Vr175M/DqJLsALwKWBbYCfgj8iWbltnuBz1fVbb2qpAQPNL+3y64uQRPc5zJsdcF2IqW/j7C64MGuLqhFxcCuXjkfOBb4GHA18FngcmB9YGngI8D9bXOnze+aFoZWF0yyJM0aBTsCL0xyGrABsEKS51XVNW2GfjzN6oIuRKRFxsCunqiqO2lWajuiquYtWZnk8cBTgaWq6sZ2X4O6esrVBdVPHDynaaGdZ/uZNJn6u3w0SNOJqwuqn5ixq+faoL4t8Gaaud8N6poWOt1AywOXDwX1odUF28z9bprVBat7jEFdvWLGrmmhDe6rVNX/2aeuXnN1QfUzA7skdbi6oPqdgV2ShmlXFzwCeD/NDIj3ADcDPwPeQRPsz6iqE21h0nRjYJekYVxdUP3MmeckLdZcXVCDxlHxkhZb7eqC/0azuuDtwJbAfjSrC25Ks7rgp4FlcHVB9QkzdkmLpXZ1wZ8DD6PpT98NmFVV5wF300wFOwf4EXAY8PGq+kNPKivNB/vYJS12XF1Qg8ymeEmLo/FWF7wZuBY4CNjT1QXVTwzskhZHri6ogWVgl7Q4cnVBDSz72CUttpKsPGx1wR1pgvqeQ6sLSv3GUfGSFltDQT3JUkmeRZO5f8igrn5mYJe0WHN1QQ0am+IlLfZcXVCDxMAuSdIAsSlekqQBYmCXJGmAGNglSRogBnZJkgaIgV3qoSRzk1yc5PIkxyZ5+AKc67AkL2jffyPJ5mPsu2OS7SZxjb8kWXWi5cP2uWM+r/W+JG+d3zpKizsDu9Rbd1XVVlW1Jc3c5Pt3NyaZMZmTVtWrq+rKMXbZEZjvwC5p+jOwS9PHL4GN2mz6zCTfAy5LMiPJoUnOT3JpktcCpPGFJFcmORlYfehESc5KMrN9v2uSC5NckuSMdsnS/YE3ta0F2ydZLclx7TXOT/KU9thVkpyW5KIkXwUy3k0k+VGSC5JckWS/Yds+2dbljCSrtWUbJjmlPeaX7TrpkibJRWCkaSDJksBuwClt0bbAllV1TRscb6uqbZIsA5yb5DTgCcCmwGOBNYArgW8NO+/QGuM7tOdauapuTfIV4I6q+kS73/do1iY/J8m6wKnAY4D3AudU1QeSPBt4UKAexSvbaywLnJ/kuKq6BVgOuLCq3pLk4PbcBwJfA/avqj8m+XfgS8DTJ/E1SsLALvXaskkubt//EvgmTRP5b6vqmrb8P4DHDfWf06wdvjGwA3BUVc0Frk/y8xHO/yTg7KFzdRc8GeYZwObN0uQAPCLJCu01ntcee3KSv0/gnt6Q5Lnt+0e1db0FuJ9mvXOA7wI/TLJ8e7/Hdq69zASuIWkUBnapt+6qqq26BW2A+1e3CHh9VZ06bL9nAeNNHZkJ7ANNt9yTq+quEeoy4ekp29XRntGe684kZwEPG2X3aq/7j+HfgaTJs49dmv5OBf5fO585STZJshxwNrB32we/JrDTCMf+GnhakvXbY1duy/8JrNDZ7zSaZnHa/bZq354NvKQt2w1YaZy6PhL4exvUN6NpMRiyBDDU6vBimib+24FrkrywvUaSPH6ca0gag4Fdmv6+QdN/fmGSy4Gv0rS2HQ/8EbgM+DLwi+EHVtVNNP3iP0xyCQ80hZ8IPHdo8BzwBmBmOzjvSh4Ynf9+YIckF9J0CfzvOHU9BVgyyaXAB4HzOtv+BWyR5AKaPvQPtOUvAV7V1u8K4DkT+E4kjcJFYCRJGiBm7JIkDRADuyRJA8TALvVQkmWSHJPk6iS/aSePGWm/s5Jc1faJX5xk9bGOT7JTZ9+Lk9ydZM9229PbSWIuT3J4+wz9wriXPZIcNInj5k2msygk2TrJZe139rl0nrMbYd91k9yRztS2SV7UjkW4IsnHO+VvTjNZ0KXtBDyP7mw7Jck/kpw0dXcmNQzs0jALK9BN0KtoRpFvBHwa+NgY+76knX52q6r621jHV9WZQ/vSDFS7EzgtyRLA4cDe7TS2fwX2XRg3UlUnVNVHF8a5ptiXaQYUbty+dh1j308DPx36kGQV4FBg56raAlgjyc7t5ouAmVX1OOAHwMc75zkUeNlCuwNpDAZ29Y3RpirNsClT27Llk3y7zcwuTfL8tvyOznEvSHJY+/6wJJ9KcibwsSTbJvlVmqlUf5Vk03a/GUk+0Tnv65PsnOT4znmfmeSHE7yt59AEWmiCwc5jZZCTPP4FwE+r6k5gFeCeqvpDu+10YOi7mZnkG8MvkGS9JL9Ps7DM5UmOTPKMJOcm+WOSbdv9Xp7kC+37F7b7XpLk7LbsId/dCNf6cpJZ7d/x+zvlH+1kw58Y7RrjSfNY4COq6tfVjBw+AthzlH33BP5MM1J/yAbAH9qnDQB+NvT9tT+m7mzLzwPWGTqoqs6gecRQmnJOUKN+8pCpSml+nD5oytR23/+hmYb1sQBJxnv+GmAT4BlVNTfJI9pzzknyDODDNP+A7wesDzyh3bYy8Hfgi0lWa//BfwXw7fa6x9BM+zrcp6rqCGBt4FqA9ny30QTfm0c45ttJ5gLHAR9qA9NEjt8b+FT7/mZgqSQzq2oWTdB/VHv8LODVo3w3GwEvbO//fJrn0J8K7AG8i4cGx4OBXarquiQrtmUjfXfDvbv9O54BnJHkccBs4LnAZlVVnfM95BrtD7BjRjgvNAvfrN2eb8jstuxB0swT8A7gmUB3hbmrgc3SdHnMbu976RGu9So6mb60KBnY1U9Gmqp0NUaeMvUZNAGNtnwiU6Ee207PCs1EK4cn2ZhmhrSlOuf9SlXN6V4vyXeAlyb5NvBk4L/a7S8a55ojZecjPYP6kjaArUAT2F9Gk22OeXyboT6WZpIb2sC4N/DpNPPOnwbMGaeOANdU1WXtOa8AzmjPdRmw3gj7nwscluT7wFDrxYjf3TB7ta0xSwJrApvTPMN/N/CNNIvdDPVTP+QaVXUVsNVoNzFKa8hI3/f7aebOv6N7SFX9Pcn/o/nxcD/wK5osvnuNlwIzgaeNVg9pKhnY1Rcy+lSlo02ZOlp5t2z4VKfdaVw/CJxZVc9ts7Ozxjnvt2kmfbmb5gfCnLbe42Xss2l+pMxu+/YfCTwk4FXVde2f/0yzYMu2NIF9vOP3Ao6vqvs65/o1sH1bv/+gaakYzz2d9/d3Pt/PCP+OVNX+aRZ0eTZwcZqZ7Mac3jbN7HhvBbZpA+hhwMPa7H5bYGeaH2sHAk8f5RqrMnbGPptOE3n7/voR9v134AVpBsetCNyf5O6q+kJVnUjzd037I2ToxyBt6867gadV1T0PPa009exjV78YbarS0aZMHT5F6lBT/I1JHpNmENlQ9j/a9a5r37+8U34asH8bROddr6qupwkQ7wEOG9q5ql7UGfDWfR3R7nICDwxeewHw8xo2a1SSJZOs2r5fCtgduHyCx+8DHDXsfPNG1NM0N3+l/bxtkiNYCJJsWFW/qaqDaZr/H8Uo313HI2h+XN2WZA2a1e5Is1DMI6vqJ8AbaTPyka5RVVeN8n1vVVX/qKobgH8meVKbvf8X8OPh9a+q7atqvapaD/gM8OGqGho/MPT9rQS8jmZmQJI8gWZWwD06gxulRc7Arn4x4lSlY0yZ+iFgpaHBVTwwj/pBNE25PwduGON6Hwc+kuRcYEan/Bs006pe2p73xZ1tRwLXVtWV83Ff3wRWSXI18Oa2fgDkgVXflgFObe/9YpofHF+fwPHr0QTU4VPNvi3J74BLgROramhVuHWBu1g4Dm0HyV1OM9/8JYz93VFVl9CMLL+CZvnZc9tNKwAntff/C+BNY1xjIv5fW5ergT/R9oWneVzvA2Md2Ppsmml3zwU+2hmIeCiwPM1KdRcnOWHogCS/BI6lGdw4O8kuE6yrNN+cUlZaSNKMCL+oqr7Z67pMRpJDge9U1aW9roukyTOwSwtBmoVN/gU8075VSb1kYJckaYDYxy5J0gAxsEuSNEAM7JIkDRADuyRJA8TALknSADGwS5I0QP4/ufWogoMitvUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 576x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "cm =  confusion_matrix(test_data['Label'],preds)\n",
    "\n",
    "plot_confusion_matrix(cm = cm, \n",
    "                      normalize    = False,\n",
    "                      target_names = ['Stock_UP','Stock_DOWN'],\n",
    "                      title        = \"Confusion Matrix of SVM\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "classifier = KNeighborsClassifier(n_neighbors=3)\n",
    "classifier.fit(train_headlines, train_data['Label'])\n",
    "preds = classifier.predict(test_headlines)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6216931216931217\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[110,  76],\n",
       "       [ 67, 125]], dtype=int64)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(accuracy_score(test_data['Label'],preds))\n",
    "confusion_matrix(test_data['Label'],preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "class1 = (test_data['Label'] != preds).sum()\n",
    "class1 = ((len(test_data['Label']) - class1)/len(test_data['Label'])) * 100\n",
    "result.append(class1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfYAAAHCCAYAAAAdAOsHAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAA2rUlEQVR4nO3deZxkVXn/8c93ZtiHfQuKCCIKQgQVUUlEFBBRo7igEBdQDJqIRon+1IjghtFojHEXN8AFwQiIooAhooKgIpssiigiExAY9p1Znt8f9/ZwaXqbnumprurPm1e9purc7VR1008955x7TqoKSZI0GGb1ugKSJGn5MbBLkjRADOySJA0QA7skSQPEwC5J0gAxsEuSNEAM7NJykmS1JN9LcluSby/DeV6R5PTlWbdeSPLDJPtPwXlflOSaJHcmecLyPr/U7wzsmnGS/H2S89rAcF0bgP52OZz6pcDGwPpVtc9kT1JV36iqZy+H+jxIkl2TVJIThpVv35afOcHzvDfJ18fbr6r2qqqjJ1ndsXwMOLiq5lbVBSPUr5I8uvP6be3PedvOZ/CZYcecleSA9vkB7T5vH7bPvCS7TsH7kZYrA7tmlCSHAJ8APkQThDcDPgu8cDmc/pHAFVW1cDmca6rcCOycZP1O2f7AFcvrAmlM5d+WRwKXTrAuhwJvAZ5RVUPH3AW8OsnmYxx6M/COJGstQz2lnjCwa8ZIsjbwfuCNVXVCVd1VVQuq6ntV9fZ2n1WSfCLJte3jE0lWabft2mZt/5LkhjYLfE277X3AYcDL25aAA4dntkk2bzPBOe3rA5L8MckdSa5K8opO+Vmd43ZO8qu2if9XSXbubDszyQeSnN2e5/QkG4zxMdwPnATs2x4/G3gZ8I1hn9V/tc3dtyf5dZKnt+XPAf618z4v6tTjiCRnA3cDj2rLXtdu/1yS/+6c/yNJzkiSEX5Os5IcmuTq9nM+Jsna7c/mTmA2cFGSP4zxPknyQeB1wC5V1f3icitwFHD4GIdfDpwDvHWsa0jTkYFdM8nTgFWBE8fY593AU4EdgO2BnYBDO9v/ClgbeDhwIPCZJOtW1eE0rQDHtU3EXx6rIknWAD4J7FVVawI7AxeOsN96wCntvusDHwdOGZZx/z3wGmAjYGXgbWNdGzgGeHX7fE+a7PfaYfv8iuYzWA/4JvDtJKtW1anD3uf2nWNeBRwErAlcPex8/wI8vv3S8nSaz27/GnlO6wPaxzOBRwFzgU9X1X1VNbfdZ/uq2nKM9/hh4OU0Qf2PI2w/AnhJkseOcY73AG9tfwZS3zCwayZZH5g/TlP5K4D3V9UNVXUj8D6agDVkQbt9QVX9ALgTGCs4jGUxsF2S1arquk5TcdfzgN9X1deqamFVHQv8Fvi7zj5fraorquoe4HiagDyqqvo5sF4b1F5NE+iH7/P1qrqpveZ/AKsw/vs8qqoubY9ZMOx8dwOvpPli8nXgTVU1b5TzvAL4eFX9saruBN4F7DvU0jFBzwZOrao/j7Sxqv4CfJ6mBWdEVXUhcDrwjqW4rtRzBnbNJDcBG4wTIB7Gg7PNq9uyJecY9sXgbpqMcqlU1V00GeUbgOuSnJJk6wnUZ6hOD++8/ssk6vM14GCarPghLRhtd8PlbfP/rTStFGM18QNcM9bGqvol8EcgNF9ARjPSz2AOzZiIidoXeGnbRTKajwB7Jtl+jH0OA/4xyV8txbWlnjKwayY5B7gX2HuMfa6lGZw1ZDMe2kw9UXcBq3dePyg4VNVpVbUHsAlNFv7FCdRnqE7/N8k6Dfka8E/AD9pseom2qfwdNH3v61bVOsBtNAEZYLQlIcdcKjLJG2ky/2uB/zfGriP9DBYC1491/mGuAHYH/inJO0esbNVNNAMpPzDaSarqt8AJNOMKpL5gYNeMUVW30WRgn0myd5LVk6yUZK8k/97udixwaJIN20Foh9E0HU/GhcAuSTZrB+69a2hDko2TvKDta7+Ppkl/0Qjn+AHwmDS36M1J8nLgccD3J1knAKrqKuAZNGMKhluTJpDeCMxJchjQHR1+PbD50ox8T/IY4IM0zfGvAv5fkh1G2f1Ymr7tLZLM5YE+/aW626Dt2tgdeHuSt4yy28dpxjdsM8ap3kczhmGdpbm+1CsGds0oVfVx4BCaAXE30jQfH0wzUhya4HMecDHwG+D8tmwy1/oRcFx7rl/z4GA8i2ZA2bU0t1Y9gyaDHn6Om4Dnt/veRJPpPr+q5k+mTsPOfVZVjdQacRrwQ5qs92qaVo5uM/vQ5Ds3JTl/vOu0XR9fBz5SVRdV1e9pMuCvDd1xMMxXaFoUfgpc1V7/TRN7Vw9WVRfRDBA8PMkbRth+O/DvNIMERzvHVW191phMHaQVLSMPSpUkSf3IjF2SpAFiYJckaYAY2CVJGiAGdkmSBoiBXZKkAbI0UzRqOZm92tq10tob9boa0kM8euM1e10FaUSXXnzB/KracKqvM3utR1YtvGfSx9c9N55WVc9ZjlVaagb2Hlhp7Y3Y9JWf7HU1pIf4ztue0esqSCPaepM1hk+tPCVq4b2ssvW+kz7+3gs+Nd7Uy1POwC5J0pAAD11NuK/Yxy5J0gAxY5ckqWviyyBMSwZ2SZK6+rwp3sAuSdIS6fuMvb9rL0mSHsSMXZKkLpviJUkaEKHvm+IN7JIkLZG+z9j7+2uJJEl6EDN2SZK6bIqXJGmA9HlTvIFdkqQl+v8+dgO7JElDXARGkiRNJ2bskiR12RQvSdKgsI9dkqTBMss+dkmSNE2YsUuSNMS54iVJGjB9frubgV2SpCX6f/Bcf9dekiQ9iIFdkqSuZPKPcU+dryS5IcklnbKPJvltkouTnJhknc62dyW5Msnvkuw5keob2CVJ6sqsyT/GdxTwnGFlPwK2q6rHA1cA7wJI8jhgX2Db9pjPJpk93gUM7JIkDVmWbH0CGXtV/RS4eVjZ6VW1sH15LrBp+/yFwLeq6r6qugq4EthpvGs4eE6SpK7eDp57LXBc+/zhNIF+yLy2bEwGdkmSlp8NkpzXeX1kVR05kQOTvBtYCHxjqGiE3Wq88xjYJUnqWrb72OdX1Y5Lf8nsDzwf2K2qhoL3POARnd02Ba4d71z2sUuStESmevDcQ6+YPAd4B/CCqrq7s+lkYN8kqyTZAtgK+OV45zNjlySpawpnnktyLLArTZP9POBwmlHwqwA/SnPtc6vqDVV1aZLjgctomujfWFWLxruGgV2SpBWkqvYbofjLY+x/BHDE0lzDwC5J0hAXgZEkaZD0/1zxBnZJkrr6fHW3/v5aIkmSHsSMXZKkLpviJUkaIH3eFG9glyRpSBw8J0nSYOnzjL2/v5ZIkqQHMWOXJKkjfZ6xG9glSWoFA7skSYMjjLwKeh+xj12SpAFixi5J0hKxKV6SpEFiYJckaYD0e2C3j12SpAFixi5JUke/Z+wGdkmShgzA7W4GdkmSWhmAUfH2sUuSNEDM2CVJ6uj3jN3ALklSh4FdkqQBYmCXJGlQDMCoeAfPSZI0QMzYJUnqsClekqQBMQj3sRvYJUnq6PfAbh+7JEkDxIxdkqSu/k7YDeySJC2R/m+KN7BLktTR74HdPnZJkgaIGbskSR39nrEb2CVJankfuyRJg6a/47qBXZKkJQZgVLyD5yRJGiBm7JIkdfR7xm5glySpw8AuSdIg6e+4bmDX9PZvL/trnvW4Dbnpzvt57sfOAmCvx/8Vb372o9lyo7m8+JM/55J5ty/Z/w3PehT77LQpixYXHzjpcn52xfxeVV0zyEqzw8PWWeVBr2+6cwG33L2QdVafw7qrz6EK7rpvETfeuaCHNdVM4OA5TWsnnDeP137xvAeVXfGXO/inoy/gV1fd/KDyR288l+ftsAl7ffQsXvul83jfi7dlVp9/81Z/WLCouPqme5c8quCOexex2sqzmLvKbP40/17+dNO93Hy3Qb0fJJn0YzowsGta+9Ufb+HWYX8M/3DDXVx1410P2Xf3bTfilAuv4/5Fi5l38z1cfdNdbL/ZOiuoplJj9ZVnsWDRYhYuLtZZbQ4337WAarctWtzTqmkCliWoT5fAblO8BsbGa6/KhVffuuT1X269l43XXrV3FdKMtNaqc7j93kUArDxnFqutPJsN5q5EATfevoB7Fxrdp7vpEqAnq28y9iTvTnJpkouTXJjkKUnekmT1SZ7vvUneNsF9z0yyY+f15kkuaZ/vmuS2JBckuTzJ4ZOpj5bdSP8rVtUIpdLUWWPV2dxx70Kg+Z2cHfjzzfdx4x0L2GSdlXtbOc0IfZGxJ3ka8HzgiVV1X5INgJWB44CvA3f3sn7Az6rq+UnWAC5M8v2q+nWP6zTj/OW2e9lknQcy9L9aZ1VuuP2+HtZIM83cVWZz34LFS5rcFy4u7rivyd7vXdAUzg4s8vvmtGbGvmJsAsyvqvsAqmo+8FLgYcCPk/wYIMl+SX6T5JIkHxk6OMlzkpyf5KIkZww/eZJ/SPLDJKstSyWr6i7g18CWy3IeTc4Zl97A83bYhJVnz2LT9VbjkRuswUV/vrXX1dIMsuaqs7m9zdahGUC3+sqzgWakfAzq/SHL8Bjv1MlXktww1Orblu3Ttkgv7rYOt9veleTKJL9LsudEqt8XGTtwOnBYkiuA/wGOq6pPJjkEeGZVzU/yMOAjwJOAW4DTk+wNnA18Edilqq5Ksl73xEkOBp4N7D30xWGykqwPPBX4wAjbDgIOApiz5kbLcpkZ5T9fsT1P2XI91l1jZc469Jn81+m/59a7F3D43o9jvbkr86UDd+Tya2/nNV88j99ffyc/uOgvnPr2p7Nw8WLee+KlLPaPqFaQAGusMpvrb79/Sdlt9yxkk7VXZvP1V6WAv9x2/6jHa/qY4oz9KODTwDGdskuAFwNfGFaPxwH7AtvSJLL/k+QxVbVorAv0RWCvqjuTPAl4OvBM4Lgk7xy225OBM6vqRoAk3wB2ARYBP62qq9pzde+RehUwjyaoj3UfykjhoVv29CQXAIuBD1fVpSO8hyOBIwFW/autDDcT9NZvXDRi+Y8uuX7E8s+d8Qc+d8YfprJK0ogKuPKGex5Sfp3BvL9M8SIwVfXTJJsPK7scRrzuC4FvtUnnVUmuBHYCzhnrGn0R2AHabyhnAmcm+Q2w/7BdRvtJhJEDMzTfknYANgWuGuPyNwHrdl6vB3RnPvlZVT1/jOMlSVpaDwfO7bye15aNqS/62JM8NslWnaIdgKuBO4A127JfAM9IskGS2cB+wE9ovtk8I8kW7bm6TfEXAK8HTm6b8kdzJvDKPPB1an/gx8v0piRJ006AZPIPYIMk53UeBy1jdYYbt8W3XzL2ucCnkqwDLASupOmv3g/4YZLrquqZSd5FE3AD/KCqvgtL+rdPSDILuAHYY+jEVXVWe9vbKUn2aAfmDXcksDVwUZICzgPeNUXvVZLUM8s80cz8qtpx/N0mZB7wiM7rTYFrxzuoLwJ7e+vYziNs+lT7GNrvm8A3Rzj+h8APh5W9t/P8NOC0Ma5/P3DwKNvOpMnoJUkDYBrd7XYy8M0kH6cZPLcV8MvxDuqLwC5J0iBIciywK02T/TzgcOBmmiR1Q5rW4wuras+qujTJ8cBlNK3VbxxvRDwY2B8kyWeAvxlW/F9V9dVe1EeStOJN8aj4/UbZdOIo+x8BHLE01zCwd1TVG3tdB0lSD2VaNcVPioFdkqRWgFl9vt5zX9zuJkmSJsaMXZKkDpviJUkaIP2+upuBXZKkIQMweM4+dkmSBogZuyRJrWau+P5O2Q3skiQtscxzxfecgV2SpI4+j+sGdkmSuvo9Y3fwnCRJA8SMXZKkIQNwu5uBXZKklqPiJUkaMH0e1+1jlyRpkJixS5LUYVO8JEkDpM/juoFdkqQl0v8Zu33skiQNEDN2SZJaze1uva7FsjGwS5K0hIvASJI0UPo8rhvYJUnq6veM3cFzkiQNEDN2SZKGuAiMJEmDw0VgJEkaMP0e2O1jlyRpgJixS5LU0ecJu4FdkqSufm+KN7BLkjRkAEbF28cuSdIAMWOXJKkV54qXJGmw9HlcN7BLktQ1q88ju4FdkqSOPo/rDp6TJGmQmLFLktRKvI9dkqSBMqu/47qBXZKkrn7P2O1jlyRpgJixS5LU0ecJu4FdkqQhoZl9rp8Z2CVJ6uj3wXP2sUuSNEDM2CVJGhIXgZEkaaD0eVw3sEuSNCT0/yIw9rFLkrSCJPlKkhuSXNIpWy/Jj5L8vv133c62dyW5Msnvkuw5kWsY2CVJ6mjmi5/cYwKOAp4zrOydwBlVtRVwRvuaJI8D9gW2bY/5bJLZ413AwC5JUkfaAXSTeYynqn4K3Dys+IXA0e3zo4G9O+Xfqqr7quoq4Epgp/GuYWCXJKm1LNn6MnTNb1xV1wG0/27Ulj8cuKaz37y2bEyjDp5L8imgRtteVW+eSG0lSeonyzh4boMk53VeH1lVR07yXCNVZNS4PGSsUfHnjbFNkiQ91Pyq2nEpj7k+ySZVdV2STYAb2vJ5wCM6+20KXDveyUYN7FV1dPd1kjWq6q6lrKwkSX2lBze7nQzsD3y4/fe7nfJvJvk48DBgK+CX451s3D72JE9Lchlweft6+ySfnVzdJUma3qZy8FySY4FzgMcmmZfkQJqAvkeS3wN7tK+pqkuB44HLgFOBN1bVovGuMZEJaj4B7EnzzYGquijJLhM4TpKkvtJMUDN156+q/UbZtNso+x8BHLE015jQqPiqumZY0bjfGCRJ0oo3kYz9miQ7A5VkZeDNtM3ykiQNlBmyCMwbgP+iuXfu/4DTgDdOZaUkSeqVPo/r4wf2qpoPvGIF1EWSpJ7r94x9IqPiH5Xke0lubCeu/26SR62IykmSpKUzkcFz36QZbr8JzX103waOncpKSZLUC0Oj4if7mA4mEthTVV+rqoXt4+tMYEo7SZL60VTex74ijDVX/Hrt0x8neSfwLZqA/nLglBVQN0mSVrjpEZ4nb6zBc7+mCeRD7/H1nW0FfGCqKiVJUi8ky7wITM+NNVf8FiuyIpIkadlN5D52kmwHPA5Ydaisqo6ZqkpJktQrfZ6wjx/YkxwO7EoT2H8A7AWcBRjYJUkDZ7oMgpusiYyKfynN5PR/qarXANsDq0xprSRJ6pFk8o/pYCKB/Z6qWgwsTLIWzQLwTlAjSdI0NJE+9vOSrAN8kWak/J1MYKF3SZL6TcjgjoofUlX/1D79fJJTgbWq6uKprZYkST0wjZrUJ2usCWqeONa2qjp/aqokSVLv9PvgubEy9v8YY1sBz1rOdZkxtn342pz94b16XQ3pIdZ98sG9roKkZTTWBDXPXJEVkSRpOpjIqPLpbEIT1EiSNBOEwW6KlyRpxpkuy69OloFdkqSOfg/s43YlpPHKJIe1rzdLstPUV02SJC2tiYwR+CzwNGC/9vUdwGemrEaSJPVIMzVsJv2YDibSFP+UqnpikgsAquqWJCtPcb0kSeqJfm+Kn0hgX5BkNs296yTZEFg8pbWSJKlHpkniPWkTaYr/JHAisFGSI2iWbP3QlNZKkiRNykTmiv9Gkl/TLN0aYO+qunzKayZJ0goWGPxFYJJsBtwNfK9bVlV/nsqKSZLUCzNh5rlTaPrXA6wKbAH8Dth2CuslSVJP9HnCPqGm+L/uvm5XfXv9lNVIkiRN2lLPPFdV5yd58lRURpKkXkoyI/rYD+m8nAU8EbhxymokSVIP9Xlcn1DGvmbn+UKaPvfvTE11JEnqrYGeoKadmGZuVb19BdVHkiQtg1EDe5I5VbWwHSwnSdLAG/T72H9J059+YZKTgW8Ddw1trKoTprhukiStcH0e1yfUx74ecBPwLB64n70AA7skabBksPvYN2pHxF/CAwF9SE1prSRJ6pHQ35F9rMA+G5gLI75DA7skSdPQWIH9uqp6/wqriSRJPdYMnut1LZbNWIG9z9+aJElLb5AD+24rrBaSJE0T6fNh8aOuTldVN6/IikiSpGW31IvASJI0qAa9j12SpJklM2OCGkmSZox+n1J21D52SZLUfwzskiS1hvrYJ/uY0DWSf05ySZJLk7ylLVsvyY+S/L79d93JvgcDuyRJHcnkH+OfO9sB/wDsBGwPPD/JVsA7gTOqaivgjPb1pBjYJUlaIsxahscEbAOcW1V3V9VC4CfAi4AXAke3+xwN7D3Zd2BglySpFaY2Y6dZWG2XJOsnWR14LvAIYOOqug6g/Xejyb4HR8VLkrT8bJDkvM7rI6vqyKEXVXV5ko8APwLuBC4CFi7PChjYJUkasuzrsc+vqh3H2qGqvgx8GSDJh4B5wPVJNqmq65JsAtww2QrYFC9JUsesZNKPiUiyUfvvZsCLgWOBk4H92132B7472fqbsUuS1BrqY59i30myPrAAeGNV3ZLkw8DxSQ4E/gzsM9mTG9glSVqBqurpI5TdxHJaVdXALklSR79PKWtglySpo8/juoFdkqQhof9Hlfd7/SVJUocZuyRJQwLp87Z4A7skSR39HdYN7JIkLdEs29rfod3ALklSR3+HdQfPSZI0UMzYJUnq6POWeAO7JEkPiKPiJUkaFE5QI0mSphUzdkmSOmyKlyRpgPR3WDewS5L0gAGYUtY+dkmSBogZuyRJrUEYFW9glySpo9+b4g3skiR19HdY7/8WB0mS1GHGLklSR5+3xBvYJUka0gye6+/IbmCXJKnDjF2SpIER0ucZu4PnJEkaIGbskiR12BQvSdKAcPCcJEmDJP2fsdvHLknSADFjlySpo98zdgO7JEkd/X67m4FdkqRWgFn9HdftY5ckaZCYsUuS1GFTvCRJA8TBc9IKttKsB/7HW7AI5nReByjg/kW9qp1mis8f/gr22mU7brz5Dnbc50MAfOgte/PcXbbj/gWLuGrefA46/Ovcduc9bLbJelx4wqFccfUNAPzyN3/izUd8q5fV1xjM2KUVaKVZsLhg0eIHyhZ0ns+ZBVUrvl6aeb72vXP5/HE/4UsfePWSsjPO/S3v+dTJLFq0mA+++YW8/bXP5tBPfheAP86bz1P3/XCvqqsJcvCctIIlsGiMwD07TeCXptrZ5/+Bm2+7+0FlZ5z7Wxa13zp/+ZurePjG6/SgZprpDOzqG0NfoleaBSvPbrLz4durfUi99uoXPo3Tzr5syevNH74+5xz7Dk7/0j/zN0/Ysoc109iyTP9NBzbFq6+Epum9aAL7nFmwsG2Knz3rwU30Uq/8vwP3ZNGixXzrB78C4C/zb+cxex3GzbfdxRO2eQTHf/wgnvjSI7jjrnt7XFM9hHPFjy3Ju5NcmuTiJBcmeUqStyRZfZLne2+St01w36OSXJXkoiRXJDkmycM729duy/7QPo5Jsna77cQke3f2/V2SQzuvv5PkxUl2TVJJ/q6z7ftJdp3M+9PYati/ixbzoO/Hs8dpppdWhFf83VN47i7bccC7j1pSdv+Chdx8210AXHD5Nfxx3ny2euRGPaqhxpNleEwHUxbYkzwNeD7wxKp6PLA7cA3wFmBSgX0S3l5V2wOPBS4Afpxk5Xbbl4E/VtWWVbUlcBXwpXbbz4Gd2/exPnAn8LTOeZ/W7gMwD3j3lL4LLVE88D/P7FkPBPlZsQlevbfHztvwLwfszkvf8gXuuXfBkvIN1p3LrHZE1uYPX59Hb7YhV82b36tqasBNZVP8JsD8qroPoKrmJ3kz8DCaADu/qp6ZZD/gX2n+Xp9SVe8ASPIc4EPA7PY8u3VPnuQfgBcDL66qe8aqSFUV8J9JXgTsleRS4EnAyzu7vR+4MsmWwNnAv7flOwPfb48LsDlwT1X9JcnWwEXASkn2qKofTeJz0lJYsAhWmt08r3pgRPzs2AyvFevofzuApz9pKzZYZy5XnvoBPvD5H/D21zybVVaew/c/dzDwwG1tf/vER/Oef3weCxctYtGi4k1HfItbbr97nCuoF5pR8dMl956cqQzspwOHJbkC+B/guKr6ZJJDgGe2gf5hwEdoguwtwOltE/jZwBeBXarqqiTrdU+c5GDg2cDeQ18cJuh8YGua5O7Cqlpyt3NVLUpyIbAtcBqwXZvd7wz8BHgUsA3whLZ+XR9sH6MG9iQHAQcBPGKzzZaiyuoa7R71BQZ1rWD7v+uoh5QdfdI5I+570hkXctIZF05thbTc9HdYn8Km+Kq6kyZgHwTcCByX5IBhuz0ZOLOqbqyqhcA3gF2ApwI/raqr2nPd3DnmVcBewEuWMqjDAz+voQHUI22v9ryXAk9s6/IL4ByaIL8zDzTDD73XnwEkefpoF66qI6tqx6raccMNNlzKakuSVpg+72Sf0sFzVbWoqs6sqsOBg4GXDNtltI9htMALcAlNc/imk6jSE4DLaYL2E5Isef/t8+3b7dAE712ANavqFuBcHgjswzN2gCOwr12S1GNTOXjusUm26hTtAFwN3AGs2Zb9AnhGkg2SzAb2o2n2Pqct36I9V7cp/gLg9cDJbVP+ROqStn9/E+DUqrqyPc+hnd0OBc5vt0ETvF9P04cOcDFN9r4ZzReDB6mq04F1ab4cSJL6VL/fxz6VGftc4OgklyW5GHgc8F7gSOCHSX5cVdcB7wJ+TBNAz6+q71bVjTRN+CckuQg4rnviqjoLeBtwSpINxqjDR9vjr6Bp9n9mVd3fbjsQeEySK5P8AXhMWzbk5zT96ue011wI3ACcV1Wj9egeweRaEiRJ00Qy+cd0kHJi7RXuSU/asc7+xXm9rob0EOs++eBeV0Ea0b0XfubXVbXjVF9nm79+Qh3z3TMnffxOW64zbj2TvBV4HU2X82+A19DcBn4cTVfzn4CXtd3AS80pZSVJ6prCwXPtRGlvBnasqu1obuneF3gncEZVbQWc0b6elL4P7Ek+085q1328ptf1kiRpFHOA1ZLMocnUrwVeCBzdbj8a2HtZTt7XquqNva6DJGkwNIn31HWWV9X/JfkY8GfgHuD0qjo9ycbtuDOq6rokk55zuO8zdkmSlptlGDjXDp7bIMl5ncdBDzp9si5Ndr4FzUysayR55fJ8C32fsUuStDwtY74+f5zBc7sDV7V3f5HkBJr5Ua5PskmbrW9CcxfWpJixS5K04vwZeGqS1dv1R3ajmRjtZGD/dp/9ge9O9gJm7JIkdU3h/ehV9Ysk/02zdslCmsnSjqSZ++X4JAfSBP99JnsNA7skSUtM/Qxy7TTrhw8rvo8me19mBnZJkjqmywxyk2UfuyRJA8SMXZKk1jRafXXSDOySJHX1eWQ3sEuS1DFdll+dLPvYJUkaIGbskiR19PuoeAO7JEkdfR7XDeySJC0xAMPiDeySJHU4eE6SJE0bZuySJLWCg+ckSRoofR7XDeySJD1In0d2+9glSRogZuySJHX0+6h4A7skSR0OnpMkaYD0eVy3j12SpEFixi5JUlefp+wGdkmSWs1U8f0d2Q3skiQNiYPnJEkaKH0e1x08J0nSIDFjlySpq89TdgO7JElLxMFzkiQNkn4fPGcfuyRJA8SMXZKkVuj7LnYDuyRJD9Lnkd3ALklSR78PnrOPXZKkAWLGLklSR7+PijewS5LU0edx3cAuSdISLgIjSdKg6e/I7uA5SZIGiBm7JEmtYFO8JEkDpc/juoFdkqSufs/Y7WOXJGmAmLFLktTR71PKGtglSerq77huYJckqavP47p97JIkDRIzdkmSWnFKWUmSBouD5yRJGiT9HdftY5ckaUVJ8tgkF3Yetyd5S5L1kvwoye/bf9ed7DUM7JIkdWQZHuOpqt9V1Q5VtQPwJOBu4ETgncAZVbUVcEb7elIM7JIkdQwNoJvMYyntBvyhqq4GXggc3ZYfDew92frbxy5J0hJZkYPn9gWObZ9vXFXXAVTVdUk2muxJzdglSWoNLdu6DBn7BknO6zwOGvE6ycrAC4BvL+/3YMYuSdLyM7+qdpzAfnsB51fV9e3r65Ns0mbrmwA3TLYCZuySJK14+/FAMzzAycD+7fP9ge9O9sRm7JIkdUz1zHNJVgf2AF7fKf4wcHySA4E/A/tM9vwGdkmSOqZ68FxV3Q2sP6zsJppR8svMpnhJkgaIGbskSUNcBEaSpMEx0RnkpjMDuyRJXX0e2e1jlyRpgJixS5LU4XrskiQNEAfPSZI0QPo8rhvYJUl6kD6P7A6ekyRpgJixS5LU4eA5SZIGxNB67P0sVdXrOsw4SW4Eru51PQbIBsD8XldCGoG/m8vPI6tqw6m+SJJTaX5ukzW/qp6zvOozGQZ29b0k51XVjr2uhzScv5vqBQfPSZI0QAzskiQNEAO7BsGRva6ANAp/N7XC2ccuSdIAMWOXJGmAGNglSRogBnZJkgaIgV2SpAFiYNeMkWSTJLsmWT/J7F7XR5qIJBsn2SaJf681Ic4VrxkhydbA8cCNbdF/ASf3rkbShL0V2Az4UJJLy1uZNA6/AWrgJdkSOA14V1XtBpwOHJJkbm9rJk3Iv9J8If1nYLuk35co0VQzsGsm2Aa4E1gZoKo+AtwLPLKXlZJGk2SrJC9Osn1VLQbeQrOYzFuAv7YrSWOxKV4DK8nmwGZV9f22f/If2n9XAeYC1/WyftJIkqxEk6XvD1yT5PPAXcCHgI8DLwOOA37Ts0pqWjOwa5A9CfhkkldU1cltUH8nTX/l1lV1e5I5VbWwt9WUGkk2Bh4PHAFcDzyF5gvoU4DPAisBewBPTPIvVXV5r+qq6cvAroFVVd9ps59PJfnnqjqpDe6vB3YFTjaoa5p5FbBXVe2W5DNAgKcBbwA2BZ4IrAH8DbBez2qpac254jVQkqwFLKiqezplfw+8C/jnqvrfJC+gGYj09ar6ao+qKi2RZKWqWtA+/xHwf1V1QJJHAm8HVgXeVlW3JlkDmF1Vt/ewyprGHDyngZFkNeBrwGuSrD5UXlXfpOmffF+SbYHvA5/CPkpNA0m2AI5I8uy26JXAGkleVlVXA58AbgW+kGTtqrrLoK6xGNg1MNos/TPAi4B9u8Ed+Dbwc2DHdpTx96rqvB5UUxpuR+CfgI8meS1wN3AWsG3bAvUH4MvAX/BODk2AfewaKFV1epL7gcMBknyrqu6uqoVJ7gI2afdb1Mt6SkMDN6vq221r0+OB59PclrkKsCdwaVUdn+S3NPMw3N3DKqtPGNg1cKrqzHYOj/cAs5NcANwPvAR4Uy/rJgEkeRjw+iQXVNVJwC+AdYFjaLLy2TRfQr+S5JqqOocmk5fGZWDXQGqD+/3AvsCzgI2A91TVT3tbM2mJucDfJXluVR2U5CXAPlX1HlgyZuQfab6UShPmqHj1rSQZb97soVHywLpVde1EjpGmSnuf+lxgUVX9KclWwL/RZOvvoxnk+a2q+nS7/6ZVNa9nFVZfMrCr73SDc5Jn0Nz687/j7du+ntUOnpNWqCTbAV+lGQy3FXAx8OGq+l2SdwPrA48AHgb8Q1Vd1rPKqq85Kl59JclGwCuTrNMWvQAYK1DPao+bDWBQVy8k2YxmGtiPVtW+wNOBRcDbkmxbVUcA3wEuoJmQxrngNWkGdvWbXYFnAC9OsgrN7/Ba0GTj3ZWvksyuqkXtl4APdL4MSCvatsC57Qj3VdrR7W8CFgKHAlTV2VX1IeBhVeUcC5o0A7v6wlDArqrjgf8BnkwzMG4WTR/6UDa+Trv/UFBfG/gucFpV3bria66ZrPNFcw3a382quq+91e0emvvXH5vkWZ3Drh92rLRUDOzqK0keXVXfAs4GdgaeA3w5yVFJfgr8b5ItOpn6ScC7q+onPau0ZqzO+I6zgK3aCWho51WY226/ALh5+DEO8tRkebub+kJVVZLnAJ9I8tKq+nqS+4ACrgW+SLPm+prt6PdVgC8Ah1fVWb2ruWaiJFvSTDbzWOA84DSaJveXJFlYVcdU1Z3tFMc70fS3S8uFo+LVF5I8GjiZZrTw2Z3yV9IMRDoP+NKwEfDeKqQVLsnWwH8DJ9KszrYzcD7N2gR30kycdCkwD9gb+Neq+k5PKquBZGDXtDZ0u1qSHWj+AL6sLV+lqu5rn78U+G1VXdI9pmeV1ozV3pf+HeDfqurYtmxz4B9oVmj7PHAvTUC/Bbiqqs72d1bLk4Fd01InoM9p+yPXBU4BPt2u1kaSvYAnVdUHe1pZiWbpVeCTNLMcHlBVdwzNm9Cu4PZp4JSq+mxPK6qBZ2DXtNMJ6s+iyWzmAb8C1qOZHvYW4CfAR4FDq+r7vaqr1JXkicAbgd8CJ1XV7zvbDgGeWVV/50RJmkqOite00wb1XYCv0IwY3oxmINJjgKOArYHn0cz9/n1vC9J0UVXn0ywdvC2wdzs2ZMgdwK/b/QzqmjJm7JqWkrwRmFVVn0qyPs2kNLsBb+WB+9bLvklNR23m/mbgMuBzNKPjvwG8sar+p5d10+AzY1fPJVkjyZPb589qbwG6DXhVkodV1U3Aj4BtgK2qBd7rq+mpzdw/SRPQP0oT1A8xqGtFMLCrp5JsQtPEfniSbwKfANakGVl8Gs1c2o8ANqaZvWtBj6oqPch4XUCdZvm1gDdV1SkrpGKa8WyKV8+09/ueCLyWZlWrLwAnVtVr2z+aTwJeRtMEfw/wn97vq15bmtUF233mVtWd7XMHzWnKGdjVE+29vd+nCdZfbm8H2pBmBrljq+rDnX1XBtauqhvtU1cvtasL7gl8r6puTfIf7fMzR9l/aM2COVW1cEXWVTOXTfHqlWcCZ7RBPcDaNItkHAf8c5LXJdkzyY+B1arqRrBPXT23K5NbXfD9ri6oFcW54tUrfwRel2RP4OXAasAOwAnAH2hWbrsf+FRV3darSkrwQPN7u+zqLJrgvohhqwu2EyndMsLqgoe5uqBWFAO7euVXwLeBjwBXAv8FXAJsAawM/BuwuG3utPld08LQ6oJJ5tCsUbArsE+S04FHAWsmeXFVXdVm6CfSrC7oQkRaYQzs6omquptmpbZjqmrJkpVJtgf+Flipqq5v9zWoq6dcXVD9xMFzmhbaebb3oMnU/9VbgzSduLqg+okZu3quDeo7AYfQzP1uUNe00OkGmgtcMhTUh1YXbDP3e2lWF6zuMQZ19YoZu6aFNrivX1V/sU9dvebqgupnBnZJ6nB1QfU7A7skDdOuLngM8D6aGRDvA+YD/wO8gybYn1FV37OFSdONgV2ShnF1QfUzZ56TNKO5uqAGjaPiJc1Y7eqCf0WzuuDtwHbAQTSrCz6WZnXB/wRWwdUF1SfM2CXNSO3qgv8LrErTn74XcF5VnQvcSzMV7ELgJOAo4N+r6oqeVFZaCvaxS5pxXF1Qg8ymeEkz0XirC84HrgHeCezt6oLqJwZ2STORqwtqYBnYJc1Eri6ogWUfu6QZK8l6w1YX3JUmqO89tLqg1G8cFS9pxhoK6klWSvJcmsz9gwZ19TMDu6QZzdUFNWhsipc047m6oAaJgV2SpAFiU7wkSQPEwC5J0gAxsEuSNEAM7JIkDRADu9RDSRYluTDJJUm+nWT1ZTjXUUle2j7/UpLHjbHvrkl2nsQ1/pRkg4mWD9vnzqW81nuTvG1p6yjNdAZ2qbfuqaodqmo7mrnJ39DdmGT2ZE5aVa+rqsvG2GVXYKkDu6Tpz8AuTR8/Ax7dZtM/TvJN4DdJZif5aJJfJbk4yesB0vh0ksuSnAJsNHSiJGcm2bF9/pwk5ye5KMkZ7ZKlbwDe2rYWPD3Jhkm+017jV0n+pj12/SSnJ7kgyReAjPcmkpyU5NdJLk1y0LBt/9HW5YwkG7ZlWyY5tT3mZ+066ZImyUVgpGkgyRxgL+DUtmgnYLuquqoNjrdV1ZOTrAKcneR04AnAY4G/BjYGLgO+Muy8Q2uM79Kea72qujnJ54E7q+pj7X7fpFmb/KwkmwGnAdsAhwNnVdX7kzwPeFCgHsVr22usBvwqyXeq6iZgDeD8qvqXJIe15z4YOBJ4Q1X9PslTgM8Cz5rExygJA7vUa6slubB9/jPgyzRN5L+sqqva8mcDjx/qP6dZO3wrYBfg2KpaBFyb5H9HOP9TgZ8Onau74MkwuwOPa5YmB2CtJGu213hxe+wpSW6ZwHt6c5IXtc8f0db1JmAxzXrnAF8HTkgyt32/3+5ce5UJXEPSKAzsUm/dU1U7dAvaAHdXtwh4U1WdNmy/5wLjTR2ZCewDTbfc06rqnhHqMuHpKdvV0XZvz3V3kjOBVUfZvdrr3jr8M5A0efaxS9PfacA/tvOZk+QxSdYAfgrs2/bBbwI8c4RjzwGekWSL9tj12vI7gDU7+51O0yxOu98O7dOfAq9oy/YC1h2nrmsDt7RBfWuaFoMhs4ChVoe/p2nivx24Ksk+7TWSZPtxriFpDAZ2afr7Ek3/+flJLgG+QNPadiLwe+A3wOeAnww/sKpupOkXPyHJRTzQFP494EVDg+eANwM7toPzLuOB0fnvA3ZJcj5Nl8Cfx6nrqcCcJBcDHwDO7Wy7C9g2ya9p+tDf35a/Ajiwrd+lwAsn8JlIGoWLwEiSNEDM2CVJGiAGdkmSBoiBXeqhJKskOS7JlUl+0U4eM9J+Kyc5MskVSX6b5CVt+SHtBDUXt5O+PLJzzKlJbk3y/WHn+lnbt35hkmuTnLSc3ssLkrxzEsctmUxnRUjypCS/aT/zT6Zzn11nn506n9FFQ7fvJVmzU35hkvlJPtFu2yzNxEIXtD+P57blj2wn37mwnbTnDcOvJy1P3u4mDZNkTlUtXEGXO5BmFPmjk+wLfAR4+Qj7vRu4oaoek2QWMDS6/QJgx3YU+j8C/945/qPA6sDruyeqqqcPPU/yHeC7y+ONVNXJwMnL41xT7HM0AwrPBX4APAf44bB9LqH5XBe2dxxclOR7VXUHsMPQTu1AwBPal4cCx1fV59LM0/8DYHPgOmDnqrqvvW//kiQnV9W1U/YONaOZsatvZJSpSjNsytS2bG6Sr7aZ2cWdDPfOznEvTXJU+/yoJB9P8mPgI23G9vM2+/p5kse2+81O8rHOed+UZLckJ3bOu0eSoT/243khcHT7/L+B3UbKIIHXAv8GUFWLq2p++/zHVXV3u8+5wKZDB1TVGTS3tY0ozQQ0zwJOal/vmORLI+y3edtK8KU0i9V8I8nuSc5O8vskO7X7HZDk0+3zfdp9L0ry07bsIZ/dCNf6XJLz2p/x+zrlH+60THxstGuMpw3Sa1XVOdWMHD4G2Hv4flV1d+fL3aqMcC9/kq1opvH92dBhwFrt87WBa9tz3V9V97Xlq+DfXU0xM3b1k4dMVUrzR/JBU6a2+76HZhrWvwZIMt791wCPAXavqkVJ1mrPuTDJ7sCHgJfQZHpbAE9ot60H3AJ8JsmG7e1lrwG+2l73OJppX4f7eFUdAzwcuAagPd9twPrA/KEdk6zTPv1Amglg/gAcXFXXDzvngTw08xzLi4Az2nvJqarzgNeNsu+jgX1o3v+vaO5D/1vgBcC/8tDgeBiwZ1X9X6f+I312w727/RnPBs5I8nhgXlvXrauqOud7yDXaL2DHjXBeaBa+eXh7viHz2rKHSDO97VeARwKvGqEVZz/guHrg1qL3Aqe3X1jWoJmoZ+hcjwBOofkc3262rqlkYFc/GWmq0g0ZecrU3YF9hw6sqolMhfrtdnpWaDKuo9usrICVOuf9/NAf+aHrJfka8MokXwWeBry63T5Ss3rXSNn58OxwDk0mfnZVHZLkEOBjwKuWnCR5JbAj8IwJvM8h+9HcIz8RV1XVb9prXUrzhaCS/IamuXm4s4GjkhzPA03VI352w7ysbY2ZA2wCPI7mHv57gS+lWexmaMzAQ65RVb+j01Q+3CitISPe81tVv6C5734bmt+FH1bVvZ1d9qXzM6D5PI+qqv9I8jTga0m2a1tYrqGZFvhhwElJ/nuEL2bScmFgV1/I6FOVjjZl6mjl3bLhU512p3H9APDjqnpRmgFtZ45z3q/STPpyL80XhIVtvcfL2OfRfEmZl2YhmLWB4QHvJuBumglpAL5Nk53TXmN3mj74Z3SafMeUZH2ahWZeNN6+re55F3deL2aEvyNV9YY2430ecGGamezGnN42zex4bwOeXFW3tN0kq7bZ/U7AbjTB9GDgWaNcYwPGztjn0emuaJ+PmT1X1eVJ7gK2A85r67o9MKeqft3Z9UCa/nqq6pwkq7b1uaFzrmvbL0ZPp+l6kZY7+3rUL0abqnS0KVOHT5E61BR/fZJt0gxAGyuorQ38X/v8gE756cAb2iC85Hpt0+q1NAOojhrauape3q63PvxxTLvLycD+7fOXAv/badodOkfRfGnYtS3ajSaLJckTaGaie0FV3cDE7QN8v5uBtuMKjhnjmAlLsmVV/aKqDqPpVngEo3x2HWvRfLm6LcnGNKvdkWbA2dpV9QPgLbQZ+UjXqKrfjfJ571BVt1bVdcAdSZ7aZu+vZoTBg0m26NTzkTRfzv7U2WU/4Nhhh/2Z5mdDm+WvCtyYZNO2+2jo9/BvgN8txccpLRUzdvWLU2mCwsU0fxTPhWbK1Lbp9oQ2WN8A7AF8kKbf+xJgEc3UqCcA76Rpyr2GZuTz3FGu9+80za+HAN1V075E0xd/cZIFNP37n263fQPYsKouW4r39WWaJtsraTL1Jd0HSS7sLI7yjna/TwBD/fjQjHyfywOro/25ql7QHv8zYGtgbpJ5wIGdhWT2BT48rC6bAfewfHy07cYIcAZwEc3nPdpnR1VdlOQCmmll/0jT1A7NnPbfbTPgAG8d4xoT8Y80X75WoxmT8ENobtejGQl/GM34gXe29VwM/NPQgMXWy4DnDjvvvwBfTPJWmpaJA9ruim2A/0hSbV0/NtStIU0Fp5SVlpM0I8IvqKov97ouk5Hko8DXquriXtdF0uQZ2KXlIM39zHcBe0y0n1uSpoKBXZKkAeLgOUmSBoiBXZKkAWJglyRpgBjYJUkaIAZ2SZIGiIFdkqQB8v8BaQoNHIoVgn4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 576x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "cm =  confusion_matrix(test_data['Label'],preds)\n",
    "\n",
    "plot_confusion_matrix(cm = cm, \n",
    "                      normalize    = False,\n",
    "                      target_names = ['Stock_UP','Stock_DOWN'],\n",
    "                      title        = \"Confusion Matrix of KNN\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAEGCAYAAABiq/5QAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAXjUlEQVR4nO3de5wlZX3n8c/XQQRFlJGGnag46k6MqGHEFkXXCwJZjEbIBhREMxg2s76SeFs1waxBdJMNGqOuGnUHVGazeMFbQJNVZ0fAGJQ4wCgg6nhBRGeZRrmKF8Bf/qinpem5nRmmTtNTn/fr1a+qes6pOr+e7vmep59T9VSqCknScNxjrguQJI2XwS9JA2PwS9LAGPySNDAGvyQNzC5zXcAo9t5771q8ePFclyFJ88pFF110bVVNzG6fF8G/ePFi1qxZM9dlSNK8kuR7m2p3qEeSBsbgl6SBMfglaWAMfkkaGINfkgbG4JekgTH4JWlgDH5JGhiDX5IGZl5cuXuXJHNdwZZ5IxxJY2aPX5IGxuCXpIEx+CVpYAx+SRqYXoM/ySuSXJ7ksiQfTLJbkoVJViVZ15Z79VmDJOnOegv+JA8EXgpMVtWjgQXAscBJwOqqWgKsbtuSpDHpe6hnF2D3JLsA9wZ+CBwJrGyPrwSO6rkGSdIMvQV/Vf0AeDNwFbAeuKGqPgvsW1Xr23PWA/tsav8ky5OsSbJmamqqrzIlaXD6HOrZi653/1Dg14D7JHnBqPtX1YqqmqyqyYmJjW4ZKUnaTn0O9RwGfLeqpqrqVuDjwJOAa5IsAmjLDT3WIEmapc/gvwp4YpJ7JwlwKHAFcA6wrD1nGXB2jzVIkmbpba6eqrowyUeBi4HbgEuAFcAewFlJTqR7czimrxokSRvrdZK2qnod8LpZzT+n6/1LkuaAV+5K0sAY/JI0MAa/JA2MwS9JA2PwS9LAGPySNDAGvyQNjMEvSQNj8EvSwBj8kjQwBr8kDYzBL0kDY/BL0sAY/JI0MAa/JA2MwS9JA9PnzdYfkWTtjK8bk7w8ycIkq5Ksa8u9+qpBkrSx3oK/qr5RVUurainwOOAW4BPAScDqqloCrG7bkqQxGddQz6HAt6vqe8CRwMrWvhI4akw1SJIYX/AfC3ywre9bVesB2nKfTe2QZHmSNUnWTE1NjalMSdr59R78SXYFngN8ZFv2q6oVVTVZVZMTExP9FCdJAzSOHv8zgYur6pq2fU2SRQBtuWEMNUiSmnEE/3HcMcwDcA6wrK0vA84eQw2SpKbX4E9yb+Bw4OMzmk8FDk+yrj12ap81SJLubJc+D15VtwAPmNX2I7qzfCRJc8ArdyVpYAx+SRoYg1+SBsbgl6SBMfglaWAMfkkaGINfkgbG4JekgTH4JWlgDH5JGhiDX5IGxuCXpIEx+CVpYHqdnVOS7vY+kLmuYMueXzv8kPb4JWlgDH5JGpi+78B1/yQfTfL1JFckOTjJwiSrkqxry736rEGSdGd99/j/J/DpqvoN4ADgCuAkYHVVLQFWt21J0pj0FvxJ9gSeCrwXoKp+UVXXA0cCK9vTVgJH9VWDJGljffb4HwZMAe9PckmS05PcB9i3qtYDtOU+m9o5yfIka5KsmZqa6rFMSRqWPoN/F+BA4N1V9VjgJ2zDsE5VraiqyaqanJiY6KtGSRqcPoP/auDqqrqwbX+U7o3gmiSLANpyQ481SJJm6S34q+r/A99P8ojWdCjwNeAcYFlrWwac3VcNkqSN9X3l7kuAM5PsCnwHeBHdm81ZSU4ErgKO6bkGSdIMvQZ/Va0FJjfx0KF9vq4kafO8cleSBsbgl6SBMfglaWAMfkkaGINfkgbG4JekgTH4JWlgDH5JGhiDX5IGZqTgT+cFSU5u2/slOajf0iRJfRi1x/8u4GDguLZ9E/B3vVQkSerVqHP1PKGqDkxyCUBVXdcmXpMkzTOj9vhvTbIAKIAkE8Ave6tKktSbUYP/7cAngH2S/BXwBeB/9FaVJKk3Iw31VNWZSS6im045wFFVdUWvlUmSejFS8CdZSHeLxA/OaLtnVd26lf2upPsg+HbgtqqabMf6MLAYuBJ4blVdtz3FS5K23ahDPRcDU8A3gXVt/btJLk7yuK3se0hVLa2q6RuynASsrqolwGq24QbskqS7btSzej4NfKKqPgOQ5LeAI4Cz6E71fMI2vOaRwNPb+krgPODPtmF/6S7J6zPXJWxWva7mugQNwKg9/snp0Aeoqs8CT62qLwH32sJ+BXw2yUVJlre2fatqfTvOemCf7ahbkrSdRu3x/zjJnwEfatvPA65rp3hu6bTOJ1fVD5PsA6xK8vVRC2tvFMsB9ttvv1F3kyRtxag9/ucDDwL+ATgb2K+1LQCeu7mdquqHbbmB7nTQg4BrkiwCaMsNm9l3RVVNVtXkxMTEiGVKkrZm1NM5rwVespmHv7WpxiT3Ae5RVTe19d8C3gCcAywDTm3Ls7e1aEnS9hv1dM4J4E+BRwG7TbdX1TO2sNu+wCeSTL/OB6rq00m+DJyV5ETgKuCY7axdkrQdRh3jP5Pu3PtnAy+m66lPbWmHqvoOcMAm2n9EdyGYJGkOjDrG/4Cqei9wa1WdX1V/ADyxx7okST0Ztcc/fYXu+iTPAn5I92GvJGmeGTX4/zLJ/YBXAu8A9gRe3ldRkqT+jBr811XVDcANwCEASZ7cW1WSpN6MOsb/jhHbJEl3c1vs8Sc5GHgSMJHkv854aE+6i7ckSfPM1oZ6dgX2aM+774z2G4Gj+ypKktSfLQZ/VZ0PnJ/kjKr63phqkiT1aNQPd++VZAXdzVN+tc9WrtyVJN0NjRr8HwHeA5xOdzctSdI8NWrw31ZV7+61EknSWIx6Oucnk/xRkkVJFk5/9VqZJKkXo/b4l7Xlq2e0FfCwHVuOJKlvo87H/9C+C5EkjcdIQz1J7p3kte3MHpIsSfLsfkuTJPVh1DH+9wO/oLuKF+Bq4C97qUiS1KtRg//hVfUm2vTMVfVTIKPsmGRBkkuSfKptL0yyKsm6ttxruyqXJG2XUYP/F0l2p/tAlyQPB34+4r4vA66YsX0SsLqqlgCr27YkaUxGDf7XAZ8GHpzkTLrA/tOt7ZTkQcCz6C78mnYksLKtrwSOGrVYSdJdN+pZPauSXEx3u8UAL6uqa0fY9W10bxAzJ3jbt6rWt+OuT7LPpnZMshxYDrDffvuNUqYkaQSjntXzu3RX7/5jVX0KuC3JUVvZ59nAhqq6aHsKq6oVVTVZVZMTExPbcwhJ0iaMPNTT7sAFQFVdTzf8syVPBp6T5ErgQ8Azkvwf4JokiwDacsO2Fi1J2n6jBv+mnre1KZ1fU1UPqqrFwLHA56rqBcA53HEl8DLg7BFrkCTtAKMG/5okb0ny8CQPS/JWYLuGcIBTgcOTrAMOb9uSpDEZda6elwB/AXy4bX8WeO2oL1JV5wHntfUfAYeOXKEkaYfaavAnWQCcXVWHjaEeSVLPtjrUU1W3A7ckud8Y6pEk9WzUoZ6fAZcmWQX8ZLqxql7aS1WSpN6MGvz/2L4kSfPcqFfurmxz9exXVd/ouSZJUo9GvXL3d4C1dPP1kGRpknN6rEuS1JNRz+M/BTgIuB6gqtYC3pVLkuahUYP/tplTNjS1o4uRJPVv1A93L0vyfGBBkiXAS4EL+itLktSXUXv8LwEeRXfzlQ8ANwAv76kmSVKPttjjT7Ib8GLg3wOXAgdX1W3jKEyS1I+t9fhXApN0of9M4M29VyRJ6tXWxvj3r6rHACR5L/Cv/ZckSerT1nr8t06vOMQjSTuHrfX4D0hyY1sPsHvbDlBVtWev1UmSdrit3UVrwbgKkSSNx6inc26zJLsl+dckX0lyeZLXt/aFSVYlWdeWe/VVgyRpY70FP905/8+oqgOApcARSZ4InASsrqolwOq2LUkak96Cvzo3t817tq8CjqQ7TZS2PKqvGiRJG+uzx0+SBUnWAhuAVVV1IbBvVa0HaMt9NrPv8iRrkqyZmprqs0xJGpReg7+qbq+qpcCDgIOSPHob9l1RVZNVNTkxMdFbjZI0NL0G/7Squh44DzgCuCbJIoC23DCOGiRJnT7P6plIcv+2vjtwGPB14BxgWXvaMuDsvmqQJG1s1GmZt8ciYGWSBXRvMGdV1aeSfBE4K8mJwFXAMT3WIEmapbfgr6qvAo/dRPuPgEP7el1J0paNZYxfknT3YfBL0sAY/JI0MAa/JA2MwS9JA2PwS9LAGPySNDAGvyQNjMEvSQNj8EvSwBj8kjQwBr8kDYzBL0kDY/BL0sAY/JI0MAa/JA1Mn7defHCSc5NckeTyJC9r7QuTrEqyri336qsGSdLG+uzx3wa8sqoeCTwR+OMk+wMnAauragmwum1Lksakt+CvqvVVdXFbvwm4AnggcCSwsj1tJXBUXzVIkjY2ljH+JIvp7r97IbBvVa2H7s0B2Gcz+yxPsibJmqmpqXGUKUmD0HvwJ9kD+Bjw8qq6cdT9qmpFVU1W1eTExER/BUrSwPQa/EnuSRf6Z1bVx1vzNUkWtccXARv6rEGSdGd9ntUT4L3AFVX1lhkPnQMsa+vLgLP7qkGStLFdejz2k4EXApcmWdva/hw4FTgryYnAVcAxPdYgSZqlt+Cvqi8A2czDh/b1upKkLfPKXUkaGINfkgbG4JekgTH4JWlgDH5JGhiDX5IGxuCXpIEx+CVpYAx+SRoYg1+SBsbgl6SBMfglaWAMfkkaGINfkgbG4JekgTH4JWlg+rz14vuSbEhy2Yy2hUlWJVnXlnv19fqSpE3rs8d/BnDErLaTgNVVtQRY3bYlSWPUW/BX1eeBH89qPhJY2dZXAkf19fqSpE0b9xj/vlW1HqAt99ncE5MsT7ImyZqpqamxFShJO7u77Ye7VbWiqiaranJiYmKuy5Gknca4g/+aJIsA2nLDmF9fkgZv3MF/DrCsrS8Dzh7z60vS4PV5OucHgS8Cj0hydZITgVOBw5OsAw5v25KkMdqlrwNX1XGbeejQvl5TkrR1d9sPdyVJ/TD4JWlgDH5JGhiDX5IGxuCXpIEx+CVpYAx+SRoYg1+SBsbgl6SBMfglaWAMfkkaGINfkgbG4JekgTH4JWlgDH5JGhiDX5IGZk6CP8kRSb6R5FtJTpqLGiRpqMYe/EkWAH8HPBPYHzguyf7jrkOShmouevwHAd+qqu9U1S+ADwFHzkEdkjRIvd1zdwseCHx/xvbVwBNmPynJcmB527w5yTfGUNso9gau3WFHS3bYoTQndujvQ07x92EnsGMz4vi79DvxkE01zkXwb+q7qI0aqlYAK/ovZ9skWVNVk3Ndh+4e/H3QbPPhd2IuhnquBh48Y/tBwA/noA5JGqS5CP4vA0uSPDTJrsCxwDlzUIckDdLYh3qq6rYkfwJ8BlgAvK+qLh93HXfB3W74SXPK3wfNdrf/nUjVRsPrkqSdmFfuStLAGPySNDDzIviT3J5kbZLLknwyyf130HFPSPLOHXSsK5Nc2upcm+RJO+K4m3idpUl+u49j686S/Lcklyf5avuZ/t8kfz3rOUuTXNHWr0zyz7MeX5vksnHWrX4kuXnG+m8nWZdkvySnJLklyT6beW4l+dsZ269KcsrYCt+EeRH8wE+ramlVPRr4MfDHc13QZhzS6lxaVReMskOSbf2AfSlg8PcsycHAs4EDq+o3gcOAU4HnzXrqscAHZmzfN8mD2zEeOY5aNV5JDgXeARxRVVe15muBV25ml58D/ynJ3uOobxTzJfhn+iLd1b8kOSjJBUkuactHtPYTknw8yafbu/KbpndO8qIk30xyPvDkGe0PSbK69e5WJ9mvtZ+R5N1Jzk3ynSRPS/K+JFckOWNLhW7lmG9Jci7wxiQPb7VelOSfk/xGe94x7a+cryT5fDv99Q3A81pPcnYIacdZBFxbVT8HqKprq+p84PokM680fy7dtCPTzuKON4fjgA+Oo1iNR5KnAKcBz6qqb8946H10/y8XbmK32+jO9HnFGEocTVXd7b+Am9tyAfARundagD2BXdr6YcDH2voJwHeA+wG7Ad+ju2hsEXAVMAHsCvwL8M62zyeBZW39D4B/aOtn0P3HDt2cQjcCj6F707wIWNqedyVwKbAWuHCEY34KWNC2VwNL2voTgM+19UuBB7b1+8/43t451z+Tnf0L2KP9LL8JvAt4Wmt/NfDWtv5E4Msz9rkS+HXggrZ9Cd1EhJfN9ffj1w75nbiVbsThN2e1nwK8CjgZeH1ru3nG4ze3rLqyZdKrgFPm8nuZLz3+3ZOsBX4ELARWtfb7AR9pY6hvBR41Y5/VVXVDVf0M+BrdnBVPAM6rqqnqJoj78IznH8wdf7L/PfAfZjz2yep+gpcC11TVpVX1S+ByYPGM500P9Uz3CLd0zI9U1e1J9gCe1L6PtcD/onuDgu6N6Ywkf0j3pqcxqaqbgcfRzRc1BXw4yQl0nYCjk9yDbphndo/+x8B1SY4FrgBuGVvR6tutwAXAiZt5/O3AsiR7zn6gqm4E/jfw0v7KG918Cf6fVtVSuvDelTvG+P87cG51Y/+/Q9e7n/bzGeu3c8fFaqNeuDDzedPH+uWs4/6SbbsIbuYxf9KW9wCurzs+G1haVY8EqKoXA6+l+2tlbZIHbMNr6S6qqtur6ryqeh3wJ8DvVdX36XpuTwN+j25oZ7YP00097jDPzuWXdEN7j0/y57MfrKrr6Tp6f7SZ/d9G96Zxn57qG9l8CX4AquoGunfMVyW5J12P/wft4RNGOMSFwNOTPKDtf8yMxy6g68EBHA98YQeUvNVjtp7Ad5McA5DOAW394VV1YVWdTPfh0YOBm4D77oDatAVJHpFkyYympXRDhtAF+luBb1fV1ZvY/RPAm+iuTtdOpKpuofvQ//gkm+r5vwX4L2yiQ1hVP6brKGzuL4axmVfBD1BVlwBfoQvUNwF/neRfGGEopKrW043HfRH4f8DFMx5+KfCiJF8FXgi8bAeUO+oxjwdOTPIVuuGj6fsT/E07RfQy4PN03/e5wP5+uNu7PYCVSb7Wfn770/3uQPc506O484e6v1JVN1XVG9twonYyLcCPAF6b5MhZj11L98Z/r83s/rd00zbPKadskKSBmXc9fknSXWPwS9LAGPySNDAGvyQNjMEvSQNj8GteSvLvknwoybfbKZf/lOTXd+RMmEnekOSwtv6UNlPn2iQPTPLR7TzmCUl+bcb26Un231E1S6PwdE7NO0lCd3Hcyqp6T2tbSndh27vbldw7+jXfQzcH0/vv4nHOA15VVWt2SGHSdrDHr/noEODW6dAHqKq1wPent5MsbjOdXty+ntTaF7WZTqfv7/CUJAvajKmXtQvmXtGee0aSo5P8Z7pL9U9OcmY79mXtOQuSvLnt99UkL2ntJyf5cjvminZF9tHAJHBme/3dk5yXZLLtc9z0BXtJ3jjje7k5yV+1WVq/lGTf1n6n2Vt7/RfXTsXg13z0aLqZUbdkA3B4VR1IN03y21v784HPtLmfDqCbgXMp3Syoj66qxwB36tVX1enAOcCrq+r4Wa+zHHgo8Njq5u0/s7W/s6oe3/762B14dlV9FFgDHN/mZPrp9EHa8M8bgWe0eh6f5Kj28H2AL1XVAXRXcP9haz8Z+I+t/Tlb+feQfsXg187qnsBpSS6lm2Jhehz9y3TTaJwCPKaqbqKbwvthSd6R5Ai6qbdHdRjwnqq6DX51OT/AIUkubK//DO48c+ymPJ47Zo69je4N5KntsV/QTeMN3Rve4rbu7K3aLga/5qPL6aZM3pJXANfQ9eon6WZ1pao+TxeoPwD+PsnvV9V17Xnn0c38evo21BJmzfiaZDe6OfyPbn9BnMadZ47d3HE259a648O4X8006+yt2l4Gv+ajzwH3aj1dAJI8nm7a7mn3A9a3+ya8kNYjTvIQYENVnQa8Fzgw3S3x7lFVHwP+AjhwG2r5LPDitFtoprsD03TIX5vufgtHz3j+5mZXvRB4WpK9kyygu3vX+Vt64c3M3ipt1bbe71Wac1VVSX4XeFuSk4Cf0c2R//IZT3sX8LE23fW53HH/g6cDr05yK92dkX6f7lae7093cxWA12xDOafT3XXrq+2Yp1XVO5OcRnfjnivphpemnQG8J8lP6W7UM/09rU/ymlZrgH+qqrO38tp/06aODt1d3L6yDXVrwDydU5IGxqEeSRoYg1+SBsbgl6SBMfglaWAMfkkaGINfkgbG4Jekgfk3RXDyJfcwmVIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = [\"RandomForest\", \"SVM\", \"KNN\"]\n",
    "h = result\n",
    "c = [\"red\", \"green\", \"orange\", \"black\"]\n",
    "plt.bar(x,h,width=0.3, color=c)\n",
    "plt.xlabel(\"Classifications\")\n",
    "plt.ylabel(\"Percentage\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}