Chapter5_BinaryClassification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "90b5eeea",
   "metadata": {},
   "source": [
    "## Minimizing Customer Attrition"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60397841",
   "metadata": {},
   "source": [
    "## Using Linear Probability Models to Find High-Risk Customers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "f6075248",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  corporation  lastmonth_activity  lastyear_activity  number_of_employees  \\\n",
      "0        abcd                  78               1024                   12   \n",
      "1        asdf                  14               2145                   20   \n",
      "2        xyzz                 182               3891                   35   \n",
      "3        acme                 101              10983                    2   \n",
      "4        qwer                   0                118                   42   \n",
      "\n",
      "   exited  \n",
      "0       1  \n",
      "1       0  \n",
      "2       0  \n",
      "3       1  \n",
      "4       1  \n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "attrition_past = pd.read_csv('https://bradfordtuckfield.com/attrition_past.csv')\n",
    "print(attrition_past.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "5247b186",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   corporation  lastmonth_activity  lastyear_activity  number_of_employees  \\\n",
      "0         abcd                  78               1024                   12   \n",
      "1         asdf                  14               2145                   20   \n",
      "2         xyzz                 182               3891                   35   \n",
      "3         acme                 101              10983                    2   \n",
      "4         qwer                   0                118                   42   \n",
      "5         tyui                 929               1992                    1   \n",
      "6         zxcv                  19                455                    8   \n",
      "7         hjkl                  94                868                    3   \n",
      "8         lmno                  81               1401                   10   \n",
      "9         qqqq                  52                650                   11   \n",
      "10        corp                1090               2452                    9   \n",
      "11        ekci                   6                 88                   90   \n",
      "12        dosk                  99                390                   99   \n",
      "13        endi                  75                800                   81   \n",
      "14        gudj                 255               1687                    2   \n",
      "15        nciw                 100               1359                    1   \n",
      "16        lsid                  68                282                   14   \n",
      "17        pwls                  71                949                   40   \n",
      "18        bqlx                 686               3782                  103   \n",
      "19        zmei                  45                655                    7   \n",
      "20        wosl                   0                 18                   21   \n",
      "21        xcvb                 189                961                   18   \n",
      "22        dfgh                  16               1028                   33   \n",
      "23        ngrd                   9                 45                    1   \n",
      "24        xful                   0                 67                   14   \n",
      "25        kshe                  48                986                   22   \n",
      "\n",
      "    exited  \n",
      "0        1  \n",
      "1        0  \n",
      "2        0  \n",
      "3        1  \n",
      "4        1  \n",
      "5        0  \n",
      "6        1  \n",
      "7        1  \n",
      "8        0  \n",
      "9        1  \n",
      "10       0  \n",
      "11       1  \n",
      "12       1  \n",
      "13       1  \n",
      "14       0  \n",
      "15       0  \n",
      "16       0  \n",
      "17       1  \n",
      "18       0  \n",
      "19       0  \n",
      "20       1  \n",
      "21       1  \n",
      "22       0  \n",
      "23       1  \n",
      "24       1  \n",
      "25       1  \n"
     ]
    }
   ],
   "source": [
    "pd.set_option('display.max_columns',6)\n",
    "# prints x # of columns; using None instead of 6 prints all columns\n",
    "print(attrition_past)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f70c6586",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5769230769230769\n"
     ]
    }
   ],
   "source": [
    "# Plotting Attrition Risk\n",
    "\n",
    "print(attrition_past['exited'].mean())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb13b5a2",
   "metadata": {},
   "source": [
    "58% of clients in the data exited their contracts in the last six months -> attrition is an issue"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "b6f4610c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjcAAAHFCAYAAAAOmtghAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA7xUlEQVR4nO3de1yUZf7/8fdwBoVRMQ4qIrYdVLRS1gOmZqam5uG37aqVimsnKtfUUjO3NLeNDmsnSyvT1LSyTW2tdSvyXFAqHlaTrFUUDxipCWiKAtfvD79MjgM66ODA7ev5eMyj5rqv+74+9zXziHf3aWzGGCMAAACL8PF2AQAAAJ5EuAEAAJZCuAEAAJZCuAEAAJZCuAEAAJZCuAEAAJZCuAEAAJZCuAEAAJZCuAEAAJZCuAGqidmzZ8tms2n9+vVlLr/tttvUqFEjp7ZGjRpp6NChFRonLS1NkyZN0pEjRy6s0PO46aabdNNNN1XKtiVp6NChLvNwPn/4wx9ks9k0fPjwMpcvXbpUkyZNKnPZM888o48//rhC45V+lrt27XK0vffee3r55ZfL7G+z2codH4Arwg1gYYsXL9YTTzxRoXXS0tL01FNPVVq4mTZtmqZNm1Yp274Qubm5+vTTTyVJ8+fP14kTJ1z6LF26VE899VSZ619IuOnVq5fS09MVHR3taDtXuElPT9c999xToTGAyxnhBrCwG264QVdeeaW3y5Ak/frrr5Kkpk2bqmnTpl6u5jdz587VqVOn1KtXLx05ckSLFi2qtLGOHz8uY4yuuOIKtW3bVoGBgW6t17ZtWzVo0KDS6gKshnADWNjZp6VKSkr09NNP65prrlFwcLBq1aqlFi1a6JVXXpEkTZo0SWPGjJEkxcXFyWazyWazaeXKlY71n3/+eV177bUKDAxURESEhgwZor179zqNe9NNNyk+Pl6rV69WYmKiQkJCNGzYMMeys09LFRYWavLkyWrSpImCgoIUHh6uzp07Ky0tzdHn9ddfV8eOHRUREaEaNWqoefPmev7553Xq1KmLmqNZs2YpMjJSc+bMUXBwsGbNmuW0fOjQoXr99dclyTEfpaeUbDabjh07pjlz5jjaS/et9NTTF198oWHDhumKK65QSEiICgsLXU5L3XTTTfr3v/+t3bt3O41RqqzTUlu3blXfvn1Vu3ZtBQUF6frrr9ecOXOc+qxcuVI2m03vv/++JkyYoHr16iksLEy33HKLtm/fflHzBlRlft4uAEDFFBcXq6ioyKXdGHPedZ9//nlNmjRJf/3rX9WxY0edOnVK33//veMU1D333KPDhw9r6tSpWrRokeO0SemRlgceeEBvvfWWhg8frttuu027du3SE088oZUrV2rDhg2qW7euY6ycnBwNGjRIY8eO1TPPPCMfn7L/X6qoqEg9evTQmjVrNHLkSN18880qKirSN998o+zsbCUmJkqSduzYoTvvvFNxcXEKCAjQ5s2b9fe//13ff/+9SyBxV1pamjIzMzVmzBiFh4fr9ttv1/z585WVlaW4uDhJ0hNPPKFjx47po48+Unp6umPd6Ohopaen6+abb1bnzp0dp//CwsKcxhg2bJh69eqld999V8eOHZO/v79LHdOmTdN9992nHTt2aPHixeete/v27UpMTFRERIReffVVhYeHa968eRo6dKh++uknjR071qn/448/rvbt2+vtt99Wfn6+xo0bp969eyszM1O+vr4VnjegyjMAqoV33nnHSDrnKzY21mmd2NhYk5SU5Hh/2223meuvv/6c47zwwgtGksnKynJqz8zMNJLMgw8+6NT+7bffGknm8ccfd7R16tTJSDLLli1z2X6nTp1Mp06dHO/nzp1rJJkZM2acewLOUFxcbE6dOmXmzp1rfH19zeHDhx3LkpKSXOahPMOGDTOSTGZmpjHGmBUrVhhJ5oknnnDq99BDD5ny/nNZo0YNpzkuVfp5DRkypNxlZ85xr169yq1bkpk4caLj/cCBA01gYKDJzs526tejRw8TEhJijhw54rQ/PXv2dOr34YcfGkkmPT29zPGA6o7TUkA1M3fuXK1bt87ldeONN5533datW2vz5s168MEH9fnnnys/P9/tcVesWCFJLndftW7dWk2aNNGyZcuc2mvXrq2bb775vNv9z3/+o6CgIMdpq/Js3LhRffr0UXh4uHx9feXv768hQ4aouLhYP/zwg9v7Uero0aP68MMPlZiYqGuvvVaS1KlTJ1155ZWaPXu2SkpKKrzNstx+++0e2c6Zli9fri5duigmJsapfejQofr111+djjBJUp8+fZzet2jRQpK0e/duj9cGVAWEG6CaadKkiRISElxedrv9vOuOHz9e//jHP/TNN9+oR48eCg8PV5cuXcq9vfxMhw4dkiSnO3xK1atXz7G8VFn9yvLzzz+rXr165Z62kqTs7Gx16NBB+/bt0yuvvKI1a9Zo3bp1jmthjh8/7tZYZ1qwYIGOHj2q/v3768iRIzpy5Ijy8vLUv39/7dmzR6mpqRXeZlncnYeKOHToULmfQ+nyM4WHhzu9L72Q+ULmDagOCDfAZcTPz0+jR4/Whg0bdPjwYb3//vvas2ePunfv7ribqTylfyBzcnJclu3fv9/pehtJThfEnssVV1yh/fv3n/NIyccff6xjx45p0aJFGjRokG688UYlJCQoICDArTHKMnPmTEnSyJEjVbt2bccrJSXFafnFcnceKiI8PLzcz0GSy2cBXG4IN8BlqlatWvrjH/+ohx56SIcPH3bcuVPe/9WXnmKaN2+eU/u6deuUmZmpLl26XFAdPXr00IkTJzR79uxy+5QGhDNvnTbGaMaMGRc0ZmZmptLT03X77bdrxYoVLq8uXbroX//6l+MIyLmOdAQGBnrkCEhFttOlSxctX77cEWZKzZ07VyEhIWrbtu1F1wNUZ9wtBVxGevfurfj4eCUkJOiKK67Q7t279fLLLys2NlZXXXWVJKl58+aSpFdeeUVJSUny9/fXNddco2uuuUb33Xefpk6dKh8fH/Xo0cNxt1RMTIxGjRp1QTXdcccdeuedd5ScnKzt27erc+fOKikp0bfffqsmTZpo4MCB6tq1qwICAnTHHXdo7NixOnHihKZPn65ffvnlgsYsPSozduxYtW7d2mV5QUGBli1bpnnz5unhhx92zMlzzz2nHj16yNfXVy1atFBAQICaN2+ulStX6pNPPlF0dLRCQ0N1zTXXVLim5s2ba9GiRZo+fbpatWolHx8fJSQklNl34sSJ+vTTT9W5c2c9+eSTqlOnjubPn69///vfev755906RQlYmrevaAbgntI7bNatW1fm8rLutjn7bqkpU6aYxMREU7duXRMQEGAaNmxo7r77brNr1y6n9caPH2/q1atnfHx8jCSzYsUKY8zpu5See+45c/XVVxt/f39Tt25dM2jQILNnzx6n9Tt16mSaNWtWZp1n3y1ljDHHjx83Tz75pLnqqqtMQECACQ8PNzfffLNJS0tz9Pnkk0/MddddZ4KCgkz9+vXNmDFjzH/+8x+n+ow5/91SJ0+eNBEREee8a6yoqMg0aNDANG/e3BhjTGFhobnnnnvMFVdcYWw2m9OdTps2bTLt27c3ISEhRpJj3871eZV1t9Thw4fNH//4R1OrVi3HGKV01t1SxhizZcsW07t3b2O3201AQIC57rrrzDvvvOPUp/RuqX/+859O7VlZWUaSS3/AKmzGuPFwDAAAgGqCa24AAIClEG4AAIClEG4AAIClEG4AAIClEG4AAIClEG4AAIClXHYP8SspKdH+/fsVGhpaKY9FBwAAnmeMUUFBwXl/i066DMPN/v37XX5JFwAAVA979uxRgwYNztnnsgs3oaGhkk5PTlhYmJerAQAA7sjPz1dMTIzj7/i5XHbhpvRUVFhYGOEGAIBqxp1LSrigGAAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWMpl94TiynL8ZLGeWbpNWQePKdjfV7c0jdTRE0WqUzNQUWFBahVbWxm7f1FuwQlFhAapdVwd+fqU/5TF4hKjb3YeUvqOQ5KM2jWuq7ZXhjutU1xitDbrsA7kn9DBgkL98utJ+djk0tfRL++4Dh49qSO/npStjH7eUFqbu/MCAMD52IwxxluDr169Wi+88IIyMjKUk5OjxYsXq1+/fudcZ9WqVRo9erS+++471atXT2PHjlVycrLbY+bn58tutysvL89jP79w79x1St2We84+Pjap5IyZjrYHaWLvpro1Ptql72dbc/TYoi068uspp/ZaIf569g/NdWt8tD7bmqOnPtmmnLwTZY5X2leSW/3KqqOylbUP55oXAMDlqyJ/v716WurYsWO67rrr9Nprr7nVPysrSz179lSHDh20ceNGPf744xoxYoQWLlxYyZWWz51gIzkHG0k6kHdCD8zboM+25ji1f7Y1R8nzNrgEG0k68uspJc/boJSl2/TAvA3lBpYz+ya72e/sOirbZ1tzytyH8uYFAAB3efXIzZlsNtt5j9yMGzdOS5YsUWZmpqMtOTlZmzdvVnp6ulvjePLIzfGTxWry5GcXvL5NUpQ9SF+Nu1m+PjYVlxi1f3a5DuSXH0ZK1/P0hxYVFqivH+tySU4JFZcY3fjc8nJD19nzAgBAtTlyU1Hp6enq1q2bU1v37t21fv16nTrleqRDkgoLC5Wfn+/08pRnlm67qPWNpJy8E1qbdViSHNfPuLOepx3IL3TUUdnWZh0+59Gks+cFAICKqFbh5sCBA4qMjHRqi4yMVFFRkQ4ePFjmOikpKbLb7Y5XTEyMx+rZdehXj2wnt+CE0z+95VKN7+443p4PAED1VK3CjXT69NWZSs+qnd1eavz48crLy3O89uzZ47FaGoWHeGQ7EaFBTv/0lks1vrvjeHs+AADVU7UKN1FRUTpw4IBTW25urvz8/BQeHl7mOoGBgQoLC3N6ecrjPZte1Po2nb47qHVcHUlS67g6igo7/x902/+9PCkqLNBRR2VrHVdH0fagcvfh7HkBAKAiqlW4adeunVJTU53avvjiCyUkJMjf3/+S1xMc4KuuTSMuaN3SP+wTezd1XDTr62PTpD7nD0z3dYy7oDHPZVKfZpfs4l1fH5sm9j69n2ePWNa8AABQEV4NN0ePHtWmTZu0adMmSadv9d60aZOys7MlnT6lNGTIEEf/5ORk7d69W6NHj1ZmZqZmzZqlmTNn6tFHH/VG+ZKkGUN+71bAOfvvdJQ9SNMHtXR5nsut8dF6Y1BL1QpxDWu1Qvz1xqCWGt+zqaYPaqloe/lHeUr7vuFmv0v9XJlb46M1fVBLRZ1VW3nzAgCAu7x6K/jKlSvVuXNnl/akpCTNnj1bQ4cO1a5du7Ry5UrHslWrVmnUqFGOh/iNGzfO6w/xk3hC8YXiCcUAAHdU5O93lXnOzaVSWeEGAABUHss+5wYAAOB8CDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSCDcAAMBSvB5upk2bpri4OAUFBalVq1Zas2bNOfvPnz9f1113nUJCQhQdHa0///nPOnTo0CWqFgAAVHVeDTcLFizQyJEjNWHCBG3cuFEdOnRQjx49lJ2dXWb/r776SkOGDNHdd9+t7777Tv/85z+1bt063XPPPZe4cgAAUFV5Ndy8+OKLuvvuu3XPPfeoSZMmevnllxUTE6Pp06eX2f+bb75Ro0aNNGLECMXFxenGG2/U/fffr/Xr11/iygEAQFXltXBz8uRJZWRkqFu3bk7t3bp1U1paWpnrJCYmau/evVq6dKmMMfrpp5/00UcfqVevXuWOU1hYqPz8fKcXAACwLq+Fm4MHD6q4uFiRkZFO7ZGRkTpw4ECZ6yQmJmr+/PkaMGCAAgICFBUVpVq1amnq1KnljpOSkiK73e54xcTEeHQ/AABA1eL1C4ptNpvTe2OMS1upbdu2acSIEXryySeVkZGhzz77TFlZWUpOTi53++PHj1deXp7jtWfPHo/WDwAAqhY/bw1ct25d+fr6uhylyc3NdTmaUyolJUXt27fXmDFjJEktWrRQjRo11KFDBz399NOKjo52WScwMFCBgYGe3wEAAFAlee3ITUBAgFq1aqXU1FSn9tTUVCUmJpa5zq+//iofH+eSfX19JZ0+4gMAAODV01KjR4/W22+/rVmzZikzM1OjRo1Sdna24zTT+PHjNWTIEEf/3r17a9GiRZo+fbp27typr7/+WiNGjFDr1q1Vr149b+0GAACoQrx2WkqSBgwYoEOHDmny5MnKyclRfHy8li5dqtjYWElSTk6O0zNvhg4dqoKCAr322mt65JFHVKtWLd1888167rnnvLULAACgirGZy+x8Tn5+vux2u/Ly8hQWFubtcgAAgBsq8vfb63dLAQAAeBLhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWArhBgAAWIrXw820adMUFxenoKAgtWrVSmvWrDln/8LCQk2YMEGxsbEKDAzUlVdeqVmzZl2iagEAQFXn583BFyxYoJEjR2ratGlq37693nzzTfXo0UPbtm1Tw4YNy1ynf//++umnnzRz5kz97ne/U25uroqKii5x5QAAoKqyGWOMtwZv06aNWrZsqenTpzvamjRpon79+iklJcWl/2effaaBAwdq586dqlOnzgWNmZ+fL7vdrry8PIWFhV1w7QAA4NKpyN9vr52WOnnypDIyMtStWzen9m7duiktLa3MdZYsWaKEhAQ9//zzql+/vq6++mo9+uijOn78eLnjFBYWKj8/3+kFAACsy2unpQ4ePKji4mJFRkY6tUdGRurAgQNlrrNz50599dVXCgoK0uLFi3Xw4EE9+OCDOnz4cLnX3aSkpOipp57yeP0AAKBq8voFxTabzem9McalrVRJSYlsNpvmz5+v1q1bq2fPnnrxxRc1e/bsco/ejB8/Xnl5eY7Xnj17PL4PAACg6vDakZu6devK19fX5ShNbm6uy9GcUtHR0apfv77sdrujrUmTJjLGaO/evbrqqqtc1gkMDFRgYKBniwcAAFWW147cBAQEqFWrVkpNTXVqT01NVWJiYpnrtG/fXvv379fRo0cdbT/88IN8fHzUoEGDSq0XAABUD149LTV69Gi9/fbbmjVrljIzMzVq1ChlZ2crOTlZ0ulTSkOGDHH0v/POOxUeHq4///nP2rZtm1avXq0xY8Zo2LBhCg4O9tZuAACAKsSrz7kZMGCADh06pMmTJysnJ0fx8fFaunSpYmNjJUk5OTnKzs529K9Zs6ZSU1P1l7/8RQkJCQoPD1f//v319NNPe2sXAABAFePV59x4A8+5AQCg+qkWz7kBAACoDBd0WurIkSNau3atcnNzVVJS4rTszGtkAAAALrUKh5tPPvlEd911l44dO6bQ0FCnZ9LYbDbCDQAA8KoKn5Z65JFHNGzYMBUUFOjIkSP65ZdfHK/Dhw9XRo0AAABuq3C42bdvn0aMGKGQkJDKqAcAAOCiVDjcdO/eXevXr6+MWgAAAC5aha+56dWrl8aMGaNt27apefPm8vf3d1rep08fjxUHAABQURV+zo2PT/kHe2w2m4qLiy+6qMrEc24AAKh+KvL3u8JHbs6+9RsAAKAq4SF+AADAUi4o3KxatUq9e/fW7373O1111VXq06eP1qxZ4+naAAAAKqzC4WbevHm65ZZbFBISohEjRmj48OEKDg5Wly5d9N5771VGjQAAAG6r8AXFTZo00X333adRo0Y5tb/44ouaMWOGMjMzPVqgp3FBMQAA1U+l/nDmzp071bt3b5f2Pn36KCsrq6KbAwAA8KgKh5uYmBgtW7bMpX3ZsmWKiYnxSFEAAAAXqsK3gj/yyCMaMWKENm3apMTERNlsNn311VeaPXu2XnnllcqoEQAAwG0VDjcPPPCAoqKiNGXKFH344YeSTl+Hs2DBAvXt29fjBQIAAFREhS8oru64oBgAgOqnUi8oBgAAqMrcOi1Vp04d/fDDD6pbt65q164tm81Wbt/Dhw97rDgAAICKcivcvPTSSwoNDXX8+7nCDQAAgDdxzQ0AAKjyKvWaG19fX+Xm5rq0Hzp0SL6+vhXdHAAAgEdVONyUd6CnsLBQAQEBF10QAADAxXD7OTevvvqqJMlms+ntt99WzZo1HcuKi4u1evVqXXvttZ6vEAAAoALcDjcvvfSSpNNHbt544w2nU1ABAQFq1KiR3njjDc9XCAAAUAFuh5vSH8Xs3LmzFi1apNq1a1daUQAAABeqwj+/sGLFisqoAwAAwCPcCjejR4/W3/72N9WoUUOjR48+Z98XX3zRI4UBAABcCLfCzcaNG3Xq1ClJ0oYNG8p9iB8P9wMAAN7GQ/wAAECVV2kP8SsqKpKfn5+2bt16UQUCAABUlgqFGz8/P8XGxqq4uLiy6gEAALgoFX5C8V//+leNHz+eX/8GAABVUoVvBX/11Vf1v//9T/Xq1VNsbKxq1KjhtHzDhg0eKw4AAKCiKhxu+vbty11RAACgyuJuKQAAUOVV2t1SktS4cWMdOnTIpf3IkSNq3LhxRTcHAADgURUON7t27SrzbqnCwkLt3bvXI0UBAABcKLevuVmyZInj3z///HPZ7XbH++LiYi1btkxxcXGerQ4AAKCC3A43/fr1c/x7UlKS0zJ/f381atRIU6ZM8VhhAAAAF8LtcFNSUiJJiouL07p161S3bt1KKwoAAOBCVfiam6eeekqhoaEu7SdPntTcuXM9UhQAAMCFqvCt4L6+vsrJyVFERIRT+6FDhxQREVHlf5qBW8EBAKh+KvVWcGNMmQ/x27t3r9NFxgAAAN7g9jU3N9xwg2w2m2w2m7p06SI/v99WLS4uVlZWlm699dZKKRIAAMBdFb5batOmTerevbtq1qzpWBYQEKBGjRrpd7/7nccLBAAAqAi3w83EiRMlSY0aNdKAAQMUFBQkScrLy9P8+fP1wgsvaPPmzVX+mhsAAGBtFb7mJikpSUFBQVq+fLkGDRqk6OhoTZ06VT179tT69esro0YAAAC3VehXwffu3avZs2dr1qxZOnbsmPr3769Tp05p4cKFatq0aWXVCAAA4Da3j9z07NlTTZs21XfffaepU6dq//79mjp1amXWBgAAUGFuH7n54osvNGLECD3wwAO66qqrKrMmAACAC+b2kZs1a9aooKBACQkJatOmjV577TX9/PPPlVkbAABAhbkdbtq1a6cZM2YoJydH999/vz744APVr19fJSUlSk1NVUFBQWXWCQAA4JYK//zCmbZv366ZM2fq3Xff1ZEjR9S1a1ctWbLEk/V5HD+/AABA9VOpP79wpmuuuUbPP/+89u7dq/fff/9iNgUAAOARFxVuSvn6+qpfv34XdNRm2rRpiouLU1BQkFq1aqU1a9a4td7XX38tPz8/XX/99RUeEwAAWJdHws2FWrBggUaOHKkJEyZo48aN6tChg3r06KHs7OxzrpeXl6chQ4aoS5cul6hSAABQXVzUNTcXq02bNmrZsqWmT5/uaGvSpIn69eunlJSUctcbOHCgrrrqKvn6+urjjz/Wpk2b3B6Ta24AAKh+Ltk1Nxfj5MmTysjIULdu3Zzau3XrprS0tHLXe+edd7Rjxw7Hb10BAACcqUI/v+BJBw8eVHFxsSIjI53aIyMjdeDAgTLX+fHHH/XYY49pzZo18vNzr/TCwkIVFhY63ufn51940QAAoMrz6jU3kmSz2ZzeG2Nc2iSpuLhYd955p5566ildffXVbm8/JSVFdrvd8YqJibnomgEAQNXltXBTt25d+fr6uhylyc3NdTmaI0kFBQVav369hg8fLj8/P/n5+Wny5MnavHmz/Pz8tHz58jLHGT9+vPLy8hyvPXv2VMr+AACAqsFrp6UCAgLUqlUrpaam6v/9v//naE9NTVXfvn1d+oeFhWnLli1ObdOmTdPy5cv10UcfKS4ursxxAgMDFRgY6NniAQBAleW1cCNJo0eP1uDBg5WQkKB27drprbfeUnZ2tpKTkyWdPuqyb98+zZ07Vz4+PoqPj3daPyIiQkFBQS7tAADg8uXVcDNgwAAdOnRIkydPVk5OjuLj47V06VLFxsZKknJycs77zBsAAIAzefU5N97Ac24AAKh+qsVzbgAAACoD4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFgK4QYAAFiK18PNtGnTFBcXp6CgILVq1Upr1qwpt++iRYvUtWtXXXHFFQoLC1O7du30+eefX8JqAQBAVefVcLNgwQKNHDlSEyZM0MaNG9WhQwf16NFD2dnZZfZfvXq1unbtqqVLlyojI0OdO3dW7969tXHjxktcOQAAqKpsxhjjrcHbtGmjli1bavr06Y62Jk2aqF+/fkpJSXFrG82aNdOAAQP05JNPutU/Pz9fdrtdeXl5CgsLu6C6AQDApVWRv99eO3Jz8uRJZWRkqFu3bk7t3bp1U1pamlvbKCkpUUFBgerUqVMZJQIAgGrIz1sDHzx4UMXFxYqMjHRqj4yM1IEDB9zaxpQpU3Ts2DH179+/3D6FhYUqLCx0vM/Pz7+wggEAQLXg9QuKbTab03tjjEtbWd5//31NmjRJCxYsUERERLn9UlJSZLfbHa+YmJiLrhkAAFRdXgs3devWla+vr8tRmtzcXJejOWdbsGCB7r77bn344Ye65ZZbztl3/PjxysvLc7z27Nlz0bUDAICqy2vhJiAgQK1atVJqaqpTe2pqqhITE8td7/3339fQoUP13nvvqVevXucdJzAwUGFhYU4vAABgXV675kaSRo8ercGDByshIUHt2rXTW2+9pezsbCUnJ0s6fdRl3759mjt3rqTTwWbIkCF65ZVX1LZtW8dRn+DgYNntdq/tBwAAqDq8Gm4GDBigQ4cOafLkycrJyVF8fLyWLl2q2NhYSVJOTo7TM2/efPNNFRUV6aGHHtJDDz3kaE9KStLs2bMvdfkAAKAK8upzbryB59wAAFD9VIvn3AAAAFQGwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUP28XYHXFJUbf7Dik9J0HJdnU7spwtW0cLl8f2znXWZt1WLkFJxQRGqTWcXVc+pfVR9I513Nnu5eCt+qoKvsPXCy+y1ULn8dvqspceD3cTJs2TS+88IJycnLUrFkzvfzyy+rQoUO5/VetWqXRo0fru+++U7169TR27FglJydfword99nWHD22aIuO/HrK0fbaiv+pVoi/nv1Dc90aH13mOk99sk05eSccbdH2IE3s3dTRv6w+tUL8JclprDPXc2e7l4K36qgq+w9cLL7LVQufx2+q0lzYjDHmko54hgULFmjw4MGaNm2a2rdvrzfffFNvv/22tm3bpoYNG7r0z8rKUnx8vO69917df//9+vrrr/Xggw/q/fff1+233+7WmPn5+bLb7crLy1NYWJind8nhs605Sp634Zx93hjU0ukD/2xrjh6Yt0FnfyClmXf6oJaSVGafspSud1/HOL21Ouuc270UXzx39q8y6vDWuICn8V2uWvg8fnMp5qIif7+9Gm7atGmjli1bavr06Y62Jk2aqF+/fkpJSXHpP27cOC1ZskSZmZmOtuTkZG3evFnp6elujXkpwk1xiVH7Z5fpQH7hOftF24P01bib5etjU3GJ0Y3PLXdKvGeySYqyB8kYc97tns3HJpWU8ymXbre0jsri7v55ug5vjQt4Gt/lqoXP4zeXai4q8vfbaxcUnzx5UhkZGerWrZtTe7du3ZSWllbmOunp6S79u3fvrvXr1+vUqVNlrlNYWKj8/HynV2Vbm3XYrQCSk3dCa7MOO9Yp74shSeb/+lc02EjlB5szt1taR2Vxd/88XYe3xgU8je9y1cLn8ZuqOBdeCzcHDx5UcXGxIiMjndojIyN14MCBMtc5cOBAmf2Liop08ODBMtdJSUmR3W53vGJiYjyzA+eQW1D+h1xe34qsUxkqe3x3t+/pOrw1LuBpfJerFj6P31TFufD6reA2m/MhKmOMS9v5+pfVXmr8+PHKy8tzvPbs2XORFZ9fRGhQhftWZJ3KUNnju7t9T9fhrXEBT+O7XLXwefymKs6F18JN3bp15evr63KUJjc31+XoTKmoqKgy+/v5+Sk8PLzMdQIDAxUWFub0qmyt4+ooKizwvP2i7b/dwt06ro6i7UEqL9bZ/q9/VFhguX3K42PTebdbWkdlcXf/PF2Ht8YFPI3vctXC5/GbqjgXXgs3AQEBatWqlVJTU53aU1NTlZiYWOY67dq1c+n/xRdfKCEhQf7+/pVWa0X5+tg0qU+z8/ab2Lup4+IqXx+bJvZuKsk1iNjO6F+6XXcCju3/Xvd2iDvvdiv7gjd398/TdXhrXMDT+C5XLXwev6mKc+HV01KjR4/W22+/rVmzZikzM1OjRo1Sdna247k148eP15AhQxz9k5OTtXv3bo0ePVqZmZmaNWuWZs6cqUcffdRbu1CuW+Oj9caglo7nz5ypdoi/y23gpetMH9RSUXbnQ3dR9iDHbXTl9akd4u8yVul643s2Pe92LwV39s9K4wKexne5auHz+E1Vmwuv3gounX6I3/PPP6+cnBzFx8frpZdeUseOHSVJQ4cO1a5du7Ry5UpH/1WrVmnUqFGOh/iNGzeuQg/xu1TPuSnFE4pd8YRi4OLwXa5a+Dx+U5lzUW2ec+MNlzrcAACAi1ctnnMDAABQGQg3AADAUgg3AADAUgg3AADAUgg3AADAUgg3AADAUgg3AADAUgg3AADAUgg3AADAUvy8XcClVvpA5vz8fC9XAgAA3FX6d9udH1a47MJNQUGBJCkmJsbLlQAAgIoqKCiQ3W4/Z5/L7relSkpKtH//foWGhspm8+wPm+Xn5ysmJkZ79uzhd6suEnPpGcyj5zCXnsNces7lNJfGGBUUFKhevXry8Tn3VTWX3ZEbHx8fNWjQoFLHCAsLs/yX7FJhLj2DefQc5tJzmEvPuVzm8nxHbEpxQTEAALAUwg0AALAUwo0HBQYGauLEiQoMDPR2KdUec+kZzKPnMJeew1x6DnNZtsvugmIAAGBtHLkBAACWQrgBAACWQrgBAACWQrgBAACWQrjxkGnTpikuLk5BQUFq1aqV1qxZ4+2SqpSUlBT9/ve/V2hoqCIiItSvXz9t377dqY8xRpMmTVK9evUUHBysm266Sd99951Tn8LCQv3lL39R3bp1VaNGDfXp00d79+69lLtS5aSkpMhms2nkyJGONubSffv27dOgQYMUHh6ukJAQXX/99crIyHAsZy7dU1RUpL/+9a+Ki4tTcHCwGjdurMmTJ6ukpMTRh7l0tXr1avXu3Vv16tWTzWbTxx9/7LTcU3P2yy+/aPDgwbLb7bLb7Ro8eLCOHDlSyXvnRQYX7YMPPjD+/v5mxowZZtu2bebhhx82NWrUMLt37/Z2aVVG9+7dzTvvvGO2bt1qNm3aZHr16mUaNmxojh496ujz7LPPmtDQULNw4UKzZcsWM2DAABMdHW3y8/MdfZKTk039+vVNamqq2bBhg+ncubO57rrrTFFRkTd2y+vWrl1rGjVqZFq0aGEefvhhRztz6Z7Dhw+b2NhYM3ToUPPtt9+arKws8+WXX5r//e9/jj7MpXuefvppEx4ebj799FOTlZVl/vnPf5qaNWual19+2dGHuXS1dOlSM2HCBLNw4UIjySxevNhpuafm7NZbbzXx8fEmLS3NpKWlmfj4eHPbbbddqt285Ag3HtC6dWuTnJzs1Hbttdeaxx57zEsVVX25ublGklm1apUxxpiSkhITFRVlnn32WUefEydOGLvdbt544w1jjDFHjhwx/v7+5oMPPnD02bdvn/Hx8TGfffbZpd2BKqCgoMBcddVVJjU11XTq1MkRbphL940bN87ceOON5S5nLt3Xq1cvM2zYMKe2P/zhD2bQoEHGGObSHWeHG0/N2bZt24wk88033zj6pKenG0nm+++/r+S98g5OS12kkydPKiMjQ926dXNq79atm9LS0rxUVdWXl5cnSapTp44kKSsrSwcOHHCax8DAQHXq1MkxjxkZGTp16pRTn3r16ik+Pv6ynOuHHnpIvXr10i233OLUzly6b8mSJUpISNCf/vQnRURE6IYbbtCMGTMcy5lL9914441atmyZfvjhB0nS5s2b9dVXX6lnz56SmMsL4ak5S09Pl91uV5s2bRx92rZtK7vdbtl5vex+ONPTDh48qOLiYkVGRjq1R0ZG6sCBA16qqmozxmj06NG68cYbFR8fL0mOuSprHnfv3u3oExAQoNq1a7v0udzm+oMPPtCGDRu0bt06l2XMpft27typ6dOna/To0Xr88ce1du1ajRgxQoGBgRoyZAhzWQHjxo1TXl6err32Wvn6+qq4uFh///vfdccdd0jie3khPDVnBw4cUEREhMv2IyIiLDuvhBsPsdlsTu+NMS5tOG348OH673//q6+++spl2YXM4+U213v27NHDDz+sL774QkFBQeX2Yy7Pr6SkRAkJCXrmmWckSTfccIO+++47TZ8+XUOGDHH0Yy7Pb8GCBZo3b57ee+89NWvWTJs2bdLIkSNVr149JSUlOfoxlxXniTkrq7+V55XTUhepbt268vX1dUm/ubm5Lmkb0l/+8hctWbJEK1asUIMGDRztUVFRknTOeYyKitLJkyf1yy+/lNvncpCRkaHc3Fy1atVKfn5+8vPz06pVq/Tqq6/Kz8/PMRfM5flFR0eradOmTm1NmjRRdna2JL6XFTFmzBg99thjGjhwoJo3b67Bgwdr1KhRSklJkcRcXghPzVlUVJR++uknl+3//PPPlp1Xws1FCggIUKtWrZSamurUnpqaqsTERC9VVfUYYzR8+HAtWrRIy5cvV1xcnNPyuLg4RUVFOc3jyZMntWrVKsc8tmrVSv7+/k59cnJytHXr1stqrrt06aItW7Zo06ZNjldCQoLuuusubdq0SY0bN2Yu3dS+fXuXRxL88MMPio2NlcT3siJ+/fVX+fg4/0nx9fV13ArOXFacp+asXbt2ysvL09q1ax19vv32W+Xl5Vl3Xr1xFbPVlN4KPnPmTLNt2zYzcuRIU6NGDbNr1y5vl1ZlPPDAA8Zut5uVK1eanJwcx+vXX3919Hn22WeN3W43ixYtMlu2bDF33HFHmbc8NmjQwHz55Zdmw4YN5uabb7b0baLuOvNuKWOYS3etXbvW+Pn5mb///e/mxx9/NPPnzzchISFm3rx5jj7MpXuSkpJM/fr1HbeCL1q0yNStW9eMHTvW0Ye5dFVQUGA2btxoNm7caCSZF1980WzcuNHxKBFPzdmtt95qWrRoYdLT0016erpp3rw5t4Lj/F5//XUTGxtrAgICTMuWLR23OOM0SWW+3nnnHUefkpISM3HiRBMVFWUCAwNNx44dzZYtW5y2c/z4cTN8+HBTp04dExwcbG677TaTnZ19ifem6jk73DCX7vvkk09MfHy8CQwMNNdee6156623nJYzl+7Jz883Dz/8sGnYsKEJCgoyjRs3NhMmTDCFhYWOPsylqxUrVpT538akpCRjjOfm7NChQ+auu+4yoaGhJjQ01Nx1113ml19+uUR7eenZjDHGO8eMAAAAPI9rbgAAgKUQbgAAgKUQbgAAgKUQbgAAgKUQbgAAgKUQbgAAgKUQbgAAgKUQbgBUe0OHDlW/fv28XUaF7dq1SzabTZs2bXJ7HZvNpo8//rjSagKsgHADVEOV+cd85cqVstlsOnLkiFv9ateurRMnTjgtW7t2rWw2m8d/cfhCwkCpm266SbNnz76o8Y8fP67atWurTp06On78eIXWLeszi4mJUU5OjuLj493eTk5Ojnr06CHp4uYDsDLCDYCLEhoaqsWLFzu1zZo1Sw0bNvRSRZVn4cKFio+PV9OmTbVo0aKL3p6vr6+ioqLk5+fn9jpRUVEKDAy86LEBKyPcABb04osvqnnz5qpRo4ZiYmL04IMP6ujRo47lu3fvVu/evVW7dm3VqFFDzZo109KlS7Vr1y517txZklS7dm3ZbDYNHTr0nGMlJSVp1qxZjvfHjx/XBx98oKSkJJe+CxcuVLNmzRQYGKhGjRppypQpTssbNWqkZ555RsOGDVNoaKgaNmyot956y7G89Nfkb7jhBtlsNt10001O6//jH/9QdHS0wsPD9dBDD+nUqVPl1j1p0iQ1bNhQgYGBqlevnkaMGHHO/ZSkmTNnatCgQRo0aJBmzpzpsvy7775Tr169FBYWptDQUHXo0EE7duzQpEmTNGfOHP3rX/9yHNFauXKl05GXkpISNWjQQG+88YbTNjds2CCbzaadO3dKcj4tVdZ8rF69Wv7+/jpw4IDTdh555BF17NjxvPsIWAHhBrAgHx8fvfrqq9q6davmzJmj5cuXa+zYsY7lDz30kAoLC7V69Wpt2bJFzz33nGrWrKmYmBgtXLhQkrR9+3bl5OTolVdeOedYgwcP1po1a5SdnS3pdIBp1KiRWrZs6dQvIyND/fv318CBA7VlyxZNmjRJTzzxhMupoilTpighIUEbN27Ugw8+qAceeEDff/+9pNOnuyTpyy+/VE5OjtPRkxUrVmjHjh1asWKF5syZo9mzZ5d7Guqjjz7SSy+9pDfffFM//vijPv74YzVv3vyc+7ljxw6lp6erf//+6t+/v9LS0hyBQ5L27dunjh07KigoSMuXL1dGRoaGDRumoqIiPfroo+rfv79uvfVW5eTkKCcnR4mJiU7b9/Hx0cCBAzV//nyn9vfee0/t2rVT48aNXWoqaz46duyoxo0b691333X0Kyoq0rx58/TnP//5nPsIWIa3f7kTQMUlJSWZvn37ut3/ww8/NOHh4Y73zZs3N5MmTSqzb+mvFJ/vF4PP7NevXz/z1FNPGWOM6dy5s3nllVfM4sWLzZn/ibnzzjtN165dnbYxZswY07RpU8f72NhYM2jQIMf7kpISExERYaZPn26MMSYrK8tIMhs3bnTaTlJSkomNjTVFRUWOtj/96U9mwIABZdY+ZcoUc/XVV5uTJ0+ecx/P9Pjjj5t+/fo53vft29dMmDDB8X78+PEmLi6u3G2W9ZmdvT8bNmwwNpvN7Nq1yxhjTHFxsalfv755/fXXHetIMosXLy5z/VLPPfecadKkieP9xx9/bGrWrGmOHj3q9v4C1RlHbgALWrFihbp27ar69esrNDRUQ4YM0aFDh3Ts2DFJ0ogRI/T000+rffv2mjhxov773/9e1HjDhg3T7NmztXPnTqWnp+uuu+5y6ZOZman27ds7tbVv314//vijiouLHW0tWrRw/LvNZlNUVJRyc3PPW0OzZs3k6+vreB8dHV3uen/60590/PhxNW7cWPfee68WL16soqKicrddXFysOXPmaNCgQY62QYMGac6cOY7aN23apA4dOsjf3/+8tZbnhhtu0LXXXqv3339fkrRq1Srl5uaqf//+FdrO0KFD9b///U/ffPONpNPXQPXv3181atS44NqA6oRwA1jM7t271bNnT8XHx2vhwoXKyMjQ66+/LkmOa1Duuece7dy5U4MHD9aWLVuUkJCgqVOnXvCYPXv21IkTJ3T33Xerd+/eCg8Pd+ljjHG5e8oY49Lv7HBgs9lUUlJy3hoqsl5MTIy2b9+u119/XcHBwXrwwQfVsWPHcq/R+fzzz7Vv3z4NGDBAfn5+8vPz08CBA7V371598cUXkqTg4ODz1uiOu+66S++9956k06ekunfvrrp161ZoGxEREerdu7feeecd5ebmaunSpRo2bJhH6gOqA8INYDHr169XUVGRpkyZorZt2+rqq6/W/v37XfrFxMQoOTlZixYt0iOPPKIZM2ZIkgICAiTJ6WjK+fj6+mrw4MFauXJluX9EmzZtqq+++sqpLS0tTVdffbXTEZdzuZDayhMcHKw+ffro1Vdf1cqVK5Wenq4tW7aU2XfmzJkaOHCgNm3a5PS66667HBcWt2jRQmvWrCk3IAUEBLhV95133qktW7YoIyNDH330UZlHwc7cplT2fNxzzz364IMP9Oabb+rKK690OWoGWJn79x8CqFLy8vJcnm9Sp04dXXnllSoqKtLUqVPVu3dvff311y534IwcOVI9evTQ1VdfrV9++UXLly9XkyZNJEmxsbGy2Wz69NNP1bNnTwUHB6tmzZrnredvf/ubxowZU+ZRG+n03Tq///3v9be//U0DBgxQenq6XnvtNU2bNs3tfY6IiFBwcLA+++wzNWjQQEFBQbLb7W6vX2r27NkqLi5WmzZtFBISonfffVfBwcGKjY116fvzzz/rk08+0ZIlS1yeR5OUlKRevXrp559/1vDhwzV16lQNHDhQ48ePl91u1zfffKPWrVvrmmuuUaNGjfT5559r+/btCg8PL7fuuLg4JSYm6u6771ZRUZH69u17QfPRvXt32e12Pf3005o8eXKF5wio1rx90Q+AiktKSjKSXF5JSUnGGGNefPFFEx0dbYKDg0337t3N3LlznS4SHj58uLnyyitNYGCgueKKK8zgwYPNwYMHHdufPHmyiYqKMjabzbHNs53vwuOzLyg2xpiPPvrING3a1Pj7+5uGDRuaF154wWl5bGyseemll5zarrvuOjNx4kTH+xkzZpiYmBjj4+NjOnXq5JiPsy/Wffjhhx3Ly6qtTZs2JiwszNSoUcO0bdvWfPnll2X2/cc//mFq1apV5oXCp06dMnXq1DFTpkwxxhizefNm061bNxMSEmJCQ0NNhw4dzI4dO4wxxuTm5pquXbuamjVrGklmxYoV5V4Q/PrrrxtJZsiQIS5j6owLisubj1JPPPGE8fX1Nfv37y9z3wCrshlTxklvAEC1d++99+qnn37SkiVLvF0KcElxWgoALCYvL0/r1q3T/Pnz9a9//cvb5QCXHOEGACymb9++Wrt2re6//3517drV2+UAlxynpQAAgKVwKzgAALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALAUwg0AALCU/w+vkmq0dGLDPQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "plt.scatter(attrition_past['lastmonth_activity'],attrition_past['exited'])\n",
    "plt.title('Historical Attrition')\n",
    "plt.xlabel('Last Month\\'s Activity')\n",
    "plt.ylabel('Attrition')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3e8d6132",
   "metadata": {},
   "source": [
    "Interpretation: Clients who were very active (>600) did not exit their contracts in the six months after their high activity was recorded."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "573ca5ef",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression()"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Confirming Relationships with Linear Regression\n",
    "\n",
    "x = attrition_past['lastmonth_activity'].values.reshape(-1,1)\n",
    "y = attrition_past['exited'].values.reshape(-1,1)\n",
    "\n",
    "from sklearn.linear_model import LinearRegression\n",
    "regressor = LinearRegression()\n",
    "regressor.fit(x, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8e1524b8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "plt.scatter(attrition_past['lastmonth_activity'],attrition_past['exited'])\n",
    "prediction = [regressor.coef_[0]*x+regressor.intercept_[0] for x in \\\n",
    "list(attrition_past['lastmonth_activity'])]\n",
    "plt.plot(attrition_past['lastmonth_activity'],  prediction, color='red')\n",
    "plt.title('Historical Attrition')\n",
    "plt.xlabel('Last Month\\'s Activity')\n",
    "plt.ylabel('Attrition')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ca268041",
   "metadata": {},
   "source": [
    "Interpretation: We estimate that a client with an activity level of 400 has a 40% attrition risk."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58ac2f57",
   "metadata": {},
   "source": [
    "###### Linear Probability Model (LPM): linear regression on binary data\n",
    "\n",
    "attrition risk = exit probability"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "fae1f481",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  corporation  lastmonth_activity  lastyear_activity  number_of_employees  \\\n",
      "0        abcd                  78               1024                   12   \n",
      "1        asdf                  14               2145                   20   \n",
      "2        xyzz                 182               3891                   35   \n",
      "3        acme                 101              10983                    2   \n",
      "4        qwer                   0                118                   42   \n",
      "\n",
      "   exited  predicted  \n",
      "0       1   0.648363  \n",
      "1       0   0.700525  \n",
      "2       0   0.563600  \n",
      "3       1   0.629618  \n",
      "4       1   0.711936  \n"
     ]
    }
   ],
   "source": [
    "attrition_past['predicted']=regressor.predict(x)\n",
    "# calculates the attrition risk for each client\n",
    "print(attrition_past.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "62413086",
   "metadata": {},
   "source": [
    "Attrition prediction: \n",
    "    (1) Learn relationships between features and target variables by using data from the past\n",
    "    (2) Use the relationships that we learned from past data to make predictions for the future"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "8eaa287e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  corporation  lastmonth_activity  lastyear_activity  number_of_employees\n",
      "0        hhtn                 166               1393                   91\n",
      "1        slfm                 824              16920                  288\n",
      "2        pryr                  68                549                   12\n",
      "3        ahva                 121               1491                   16\n",
      "4        dmai                   4                 94                    2\n"
     ]
    }
   ],
   "source": [
    "# Predicting the Future\n",
    "\n",
    "attrition_future=pd.read_csv('http://bradfordtuckfield.com/attrition2.csv')\n",
    "# new data, which was generated today\n",
    "print(attrition_future.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "198496d3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  corporation  lastmonth_activity  lastyear_activity  number_of_employees  \\\n",
      "0        hhtn                 166               1393                   91   \n",
      "1        slfm                 824              16920                  288   \n",
      "2        pryr                  68                549                   12   \n",
      "3        ahva                 121               1491                   16   \n",
      "4        dmai                   4                 94                    2   \n",
      "\n",
      "   predicted  \n",
      "0   0.648363  \n",
      "1   0.700525  \n",
      "2   0.563600  \n",
      "3   0.629618  \n",
      "4   0.711936  \n"
     ]
    }
   ],
   "source": [
    "X=attrition_future['lastmonth_activity'].values.reshape(-1,1)\n",
    "attrition_future['predicted']=regressor.predict(x)\n",
    "print(attrition_future.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "938fa429",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   corporation  lastmonth_activity  lastyear_activity  number_of_employees  \\\n",
      "4         dmai                   4                 94                    2   \n",
      "20        wlcj                 158               2813                    6   \n",
      "24        pian                   0                 12                   19   \n",
      "11        ejdc                  95               1005                   61   \n",
      "23        hlly                  45                468                    3   \n",
      "\n",
      "    predicted  \n",
      "4    0.711936  \n",
      "20   0.711936  \n",
      "24   0.711936  \n",
      "11   0.707045  \n",
      "23   0.704600  \n"
     ]
    }
   ],
   "source": [
    "# Making Business Recommendations\n",
    "\n",
    "print(attrition_future.nlargest(5, 'predicted'))\n",
    "# specifies number of clients n that have a time and bandwith that needs focus\n",
    "# creats a list of top n highest-risk clients"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "8292f156",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['pian', 'dmai', 'wlcj', 'ejdc', 'hlly', 'slfm', 'ynus', 'frgv', 'dwgt', 'hlpd', 'oscr', 'cred', 'angl', 'pevc', 'hhtn', 'whsh', 'grce', 'mike', 'xkcd', 'ahva', 'pryr', 'bass', 'erin', 'zamk', 'skct', 'kdic']\n"
     ]
    }
   ],
   "source": [
    "print(list(attrition_future.sort_values(by='predicted',ascending=False).loc[:,'corporation']))\n",
    "# sorts clients from highest to lowest attrition risk"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "d479881f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['slfm', 'dmai', 'ejdc', 'wlcj', 'hlly', 'pian']\n"
     ]
    }
   ],
   "source": [
    "print(list(attrition_future.loc[attrition_future['predicted']>0.7,'corporation']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "35f36ba9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Measuring Prediction Accuracy\n",
    "\n",
    "themedian=attrition_past['predicted'].median() # median of predicted column\n",
    "prediction=list(1*(attrition_past['predicted']>themedian))\n",
    "# 0 when LPM predicts below-medium probability, 1 when above-median probability\n",
    "actual=list(attrition_past['exited'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "deef4068",
   "metadata": {},
   "source": [
    "Every confusion matrix has the following structure: \\\n",
    "[[true positives    false positives] \\\n",
    " [false negatives   true negatives]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "8f718782",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[7 6]\n",
      " [4 9]]\n"
     ]
    }
   ],
   "source": [
    "# confusion matrix - a standard way to measure accuracy in classification models\n",
    "from sklearn.metrics import confusion_matrix\n",
    "print(confusion_matrix(prediction,actual))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3dfc14d3",
   "metadata": {},
   "source": [
    "Seven true-positive classifications: for seven corporations, our model predicted above-median exit probability (high attrition risk), and those seven corporation did exit.\n",
    "\n",
    "Six false-positives: Six cases in which we predicted above-median exit probability but the corporation didn't exit.\n",
    "\n",
    "Four false-negatives: Four cases in which we predicted below-median exit probability but the corporation did exit.\n",
    "\n",
    "Nine true-negatives: Nine cases in which we predicted below-median exit probability for clients that did not exit."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "46683525",
   "metadata": {},
   "source": [
    "Precision: true positives / (true positives + false positives)\n",
    "- Out of everything we thought was positive, how many times was it actually positive? \n",
    "    - (In our case, positive refers to attrition—out of all the times we thought a client was at high risk of leaving, how many times did they actually leave?)\n",
    "\n",
    "Recall (aka sensitivity): true positives / (true positives + false negatives)\n",
    "- Out of all the actually positive cases, how many did we think were positive? \n",
    "    - (In other words, out of all the clients who actually exited their contracts, how many did we predict were at high attrition risk?) \n",
    "    \n",
    "If false positives are high, precision will be low. <br>\n",
    "If false negatives are high, recall will be low. <br>\n",
    "Ideally, both will be as high as possible."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "e7f35cd9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5384615384615384\n",
      "0.6363636363636364\n"
     ]
    }
   ],
   "source": [
    "conf_mat = confusion_matrix(prediction,actual)\n",
    "precision = conf_mat[0][0]/(conf_mat[0][0]+conf_mat[0][1])\n",
    "recall = conf_mat[0][0]/(conf_mat[0][0]+conf_mat[1][0])\n",
    "\n",
    "print(precision)\n",
    "print(recall)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "5b257192",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression()"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Using Multivariate LPMs\n",
    "\n",
    "x3 = attrition_past.loc[:,['lastmonth_activity', 'lastyear_activity','number_of_employees']].values.reshape(-1,3)\n",
    "y = attrition_past['exited'].values.reshape(-1,1)\n",
    "regressor_multi = LinearRegression()\n",
    "regressor_multi.fit(x3, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "016af2b4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   corporation  lastmonth_activity  lastyear_activity  number_of_employees  \\\n",
      "11        ejdc                  95               1005                   61   \n",
      "12        mike                   0                  0                   49   \n",
      "13        pevc                   4                  6                 1686   \n",
      "4         dmai                   4                 94                    2   \n",
      "22        ynus                   9                 90                   12   \n",
      "\n",
      "    predicted  predicted_multi  \n",
      "11   0.707045         0.870000  \n",
      "12   0.631248         0.815677  \n",
      "13   0.650809         0.788110  \n",
      "4    0.711936         0.755625  \n",
      "22   0.698895         0.715362  \n"
     ]
    }
   ],
   "source": [
    "# multivariate linear probability model\n",
    "attrition_future['predicted_multi']=regressor_multi.predict(x3)\n",
    "\n",
    "print(attrition_future.nlargest(5,'predicted_multi'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "31cac6c2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['ejdc', 'mike', 'pevc', 'dmai', 'ynus', 'wlcj', 'angl', 'pian', 'slfm', 'hlpd', 'frgv', 'hlly', 'oscr', 'cred', 'dwgt', 'hhtn', 'whsh', 'grce', 'pryr', 'xkcd', 'bass', 'ahva', 'erin', 'zamk', 'skct', 'kdic']\n"
     ]
    }
   ],
   "source": [
    "print(list(attrition_future.sort_values(by='predicted_multi',ascending=False).loc[:,'corporation']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "d39bb029",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating New Metrics\n",
    "\n",
    "# derived feature (or engineered feature) - a feature or metric created by transforming and combing existing variables\n",
    "attrition_future['activity_per_employee']=attrition_future.loc[:,'lastmonth_activity']/attrition_future.loc[:,'number_of_employees']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "55cdf9bd",
   "metadata": {},
   "outputs": [],
   "source": [
    "attrition_past['activity_per_employee']=attrition_past.loc[:,'lastmonth_activity']/attrition_past.loc[:,'number_of_employees']\n",
    "x = attrition_past.loc[:,['activity_per_employee','lastmonth_activity','lastyear_activity', 'number_of_employees']].values.reshape(-1,4)\n",
    "y = attrition_past['exited'].values.reshape(-1,1)\n",
    "\n",
    "regressor_derived= LinearRegression()\n",
    "regressor_derived.fit(x, y)\n",
    "attrition_past['predicted3']=regressor_derived.predict(x)\n",
    "\n",
    "x = attrition_future.loc[:,['activity_per_employee','lastmonth_activity','lastyear_activity', 'number_of_employees']].values.reshape(-1,4)\n",
    "attrition_future['predicted3']=regressor_derived.predict(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "f8eb1e0f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['pevc', 'bass', 'frgv', 'hlpd', 'angl', 'oscr', 'zamk', 'whsh', 'mike', 'hhtn', 'ejdc', 'grce', 'pian', 'ynus', 'dmai', 'kdic', 'erin', 'slfm', 'dwgt', 'pryr', 'hlly', 'xkcd', 'skct', 'ahva', 'wlcj', 'cred']\n"
     ]
    }
   ],
   "source": [
    "print(list(attrition_future.sort_values(by='predicted3',ascending=False).loc[:,'corporation']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "d78fdda3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 9  4]\n",
      " [ 2 11]]\n"
     ]
    }
   ],
   "source": [
    "themedian=attrition_past['predicted3'].median()\n",
    "prediction=list(1*(attrition_past['predicted3']>themedian))\n",
    "actual=list(attrition_past['exited'])\n",
    "\n",
    "print(confusion_matrix(prediction,actual))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "d08d3b00",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6923076923076923\n",
      "0.8181818181818182\n"
     ]
    }
   ],
   "source": [
    "conf_mat = confusion_matrix(prediction,actual)\n",
    "precision = conf_mat[0][0]/(conf_mat[0][0]+conf_mat[0][1])\n",
    "recall = conf_mat[0][0]/(conf_mat[0][0]+conf_mat[1][0])\n",
    "\n",
    "print(precision)\n",
    "print(recall)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "014ca452",
   "metadata": {},
   "source": [
    "###### Considering the Weaknesses of LPMS\n",
    "\n",
    "- LPMs don’t fit the points of a dataset well: they pass through the middle of the points and get close to only a few points\n",
    "- no reasonable way to interpret a negative probability; it’s just nonsense that our model has output\n",
    "\n",
    "Any non-horizontal regression line will make predictions that are below 0 percent or above 100 percent for certain values."
   ]
  },
  {
   "attachments": {
    "Screenshot%202023-06-19%20at%207.13.58%20PM.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "id": "ab5cd376",
   "metadata": {},
   "source": [
    "## Predicting Binary Outcomes with Logistic Regression\n",
    "\n",
    "![Screenshot%202023-06-19%20at%207.13.58%20PM.png](attachment:Screenshot%202023-06-19%20at%207.13.58%20PM.png)\n",
    "\n",
    "linear regression: β0 + β1·x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "34cb6bb0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Drawing Logistic Curves\n",
    "\n",
    "from matplotlib import pyplot as plt\n",
    "import numpy as np\n",
    "import math\n",
    "x = np.arange(-5, 5, 0.05)\n",
    "y = (1/(1+np.exp(-1-2*x)))\n",
    "plt.plot(x,y)\n",
    "plt.xlabel(\"X\")\n",
    "plt.ylabel(\"Value of Logistic Function\")\n",
    "plt.title('A Logistic Curve')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "c7f2e889",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# changing negatives and positives changes direction\n",
    "\n",
    "from matplotlib import pyplot as plt\n",
    "import numpy as np\n",
    "import math\n",
    "x = np.arange(-5, 5, 0.05)\n",
    "y = (1/(1+np.exp(1+2*x)))\n",
    "plt.plot(x,y)\n",
    "plt.xlabel(\"X\")\n",
    "plt.ylabel(\"Value of Logistic Function\")\n",
    "plt.title('A Logistic Curve')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "703cdd47",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression(random_state=0, solver='liblinear')"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Fitting the Logistic Function to Our Data\n",
    "\n",
    "# logistic regression: fitting a logistic curve to binary data\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "model = LogisticRegression(solver='liblinear', random_state=0)\n",
    "x = attrition_past['lastmonth_activity'].values.reshape(-1,1)\n",
    "y = attrition_past['exited']\n",
    "model.fit(x, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "44e50c97",
   "metadata": {},
   "outputs": [],
   "source": [
    "attrition_past['logisticprediction']=model.predict_proba(x)[:,1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "7f6edc29",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.scatter(attrition_past['lastmonth_activity'],attrition_past['exited'], color='blue')\n",
    "attrition_past.sort_values('lastmonth_activity').plot('lastmonth_activity',\\\n",
    "'logisticprediction',ls='--', ax=fig.axes,color='red')\n",
    "plt.title('Logistic Regression for Attrition Predictions')\n",
    "plt.xlabel('Last Month\\'s Activity')\n",
    "plt.ylabel('Attrition (1=Exited)')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7b0cda96",
   "metadata": {},
   "source": [
    "A regression that never predicts above 100 percent or below 0 percent probability and gets very close to some of the points in our strange “cloud.” <br>\n",
    "We’ve resolved the weaknesses of LPMs with this new method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "51e21125",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['whsh', 'pian', 'mike', 'bass', 'pevc', 'dmai', 'ynus', 'kdic', 'hlpd', 'angl', 'erin', 'oscr', 'grce', 'zamk', 'hlly', 'xkcd', 'dwgt', 'pryr', 'skct', 'frgv', 'ejdc', 'ahva', 'wlcj', 'hhtn', 'slfm', 'cred']\n"
     ]
    }
   ],
   "source": [
    "x = attrition_future['lastmonth_activity'].values.reshape(-1,1)\n",
    "attrition_future['logisticprediction']=model.predict_proba(x)[:,1]\n",
    "print(list(attrition_future.sort_values(by='logisticprediction',\\\n",
    "ascending=False).loc[:,'corporation']))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9263c802",
   "metadata": {},
   "source": [
    "## Applications of Binary Classification\n",
    "\n",
    "- Logistic regressions and LPMs are commonly used to predict binary outcomes.\n",
    "- LPMs and logistic regressions are both monotonic: they express a trend that moves in only one direction\n",
    "    - They will be unable to predict an up down up trend"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ec903046",
   "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}