GradientBoostingRegression.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv(\"mpg.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data[data.horsepower != \"?\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "mpg             float64\n",
       "cylinders         int64\n",
       "displacement    float64\n",
       "horsepower       object\n",
       "weight            int64\n",
       "acceleration    float64\n",
       "model_year        int64\n",
       "origin            int64\n",
       "name             object\n",
       "dtype: object"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.horsepower = data.horsepower.astype(\"float\")      #object type has been changed to float"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import mean_squared_error, r2_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "df=data\n",
    "training_features = [ 'acceleration','origin','displacement','model_year','horsepower']\n",
    "target = 'mpg'\n",
    "\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(df[training_features],\n",
    "                                                    df[target],\n",
    "                                                     test_size=0.10, random_state=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import GradientBoostingRegressor\n",
    "model = GradientBoostingRegressor(n_estimators=100)\n",
    "\n",
    "#n_estimators here mean weak learners that will be clubbed into one strong model\n",
    "# but instead of using the normal approach the resultant storng model will be arrived at\n",
    "# by gradient updatation at each iteration."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradientBoostingRegressor()"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train,Y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean squared error: 2.47\n",
      "R Square score: 0.97\n"
     ]
    }
   ],
   "source": [
    "predicted=model.predict(X_test)\n",
    "print(\"Mean squared error: %.2f\"\n",
    "              % mean_squared_error(Y_test, predicted))\n",
    "print('R Square score: %.2f' % r2_score(Y_test, predicted))\n",
    "\n",
    "#we should get a better rsquare value than adaboost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(167.4, 190.26, 'X[2] <= 190.5\\nmse = 73.091\\nsamples = 352\\nvalue = 25.09'),\n",
       " Text(83.7, 135.9, 'X[3] <= 77.5\\nmse = 44.217\\nsamples = 222\\nvalue = 29.742'),\n",
       " Text(41.85, 81.53999999999999, 'X[4] <= 96.5\\nmse = 18.994\\nsamples = 91\\nvalue = 25.011'),\n",
       " Text(20.925, 27.180000000000007, 'mse = 14.772\\nsamples = 76\\nvalue = 26.105'),\n",
       " Text(62.775000000000006, 27.180000000000007, 'mse = 3.582\\nsamples = 15\\nvalue = 19.467'),\n",
       " Text(125.55000000000001, 81.53999999999999, 'X[2] <= 97.5\\nmse = 35.385\\nsamples = 131\\nvalue = 33.029'),\n",
       " Text(104.625, 27.180000000000007, 'mse = 37.117\\nsamples = 41\\nvalue = 38.066'),\n",
       " Text(146.475, 27.180000000000007, 'mse = 17.774\\nsamples = 90\\nvalue = 30.734'),\n",
       " Text(251.10000000000002, 135.9, 'X[4] <= 127.0\\nmse = 22.32\\nsamples = 130\\nvalue = 17.145'),\n",
       " Text(209.25, 81.53999999999999, 'X[3] <= 81.5\\nmse = 16.746\\nsamples = 65\\nvalue = 20.392'),\n",
       " Text(188.32500000000002, 27.180000000000007, 'mse = 7.327\\nsamples = 57\\nvalue = 19.605'),\n",
       " Text(230.175, 27.180000000000007, 'mse = 48.0\\nsamples = 8\\nvalue = 26.0'),\n",
       " Text(292.95, 81.53999999999999, 'X[4] <= 186.5\\nmse = 6.809\\nsamples = 65\\nvalue = 13.898'),\n",
       " Text(272.02500000000003, 27.180000000000007, 'mse = 4.858\\nsamples = 54\\nvalue = 14.637'),\n",
       " Text(313.875, 27.180000000000007, 'mse = 0.562\\nsamples = 11\\nvalue = 10.273')]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.tree import plot_tree\n",
    "plot_tree(model.estimators_[2],filled=True)\n",
    "\n",
    "#reduce depth to get a clear picture with values\n",
    "# model.estimators_[i] here i can be 0 to 21 since we have 22 estimators"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}