Modules/2-Supervised Learning/notebooks/supervised_learning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline \n",
    "\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LinearRegression, Ridge, Lasso\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "from sklearn.neighbors import KNeighborsRegressor\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.ensemble import RandomForestRegressor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load data\n",
    "\n",
    "- Dataset Description: https://www4.stat.ncsu.edu/~boos/var.select/diabetes.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [],
   "source": [
    "diabetes_df = pd.read_csv('../../../Datasets/diabetes.txt', delimiter='\\t')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['AGE', 'SEX', 'BMI', 'BP', 'S1', 'S2', 'S3', 'S4', 'S5', 'S6', 'Y'], dtype='object')"
      ]
     },
     "execution_count": 181,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diabetes_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "diabetes_df.plot(kind='scatter', x='AGE', y='BP')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(442, 11)"
      ]
     },
     "execution_count": 206,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diabetes_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(397, 9) (397, 1) (45, 9) (45, 1)\n"
     ]
    }
   ],
   "source": [
    "# let's predict BP from AGE\n",
    "x_train, x_test, y_train, y_test = train_test_split(diabetes_df[['AGE', 'SEX', 'BMI', 'S1', 'S2', 'S3', 'S4', 'S5', 'S6']], diabetes_df[['BP']], test_size=0.1, random_state=2018)\n",
    "print(x_train.shape, y_train.shape, x_test.shape, y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Simple Linear Regression\n",
    "\n",
    "- link: https://en.wikipedia.org/wiki/Simple_linear_regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Coefficients: \n",
      " [[ 0.18244846  4.53119284  0.76888844  0.1229513  -0.08345719 -0.15997197\n",
      "  -2.15921383  3.02465792  0.19040928]]\n",
      "Train stats: \n",
      "\tmean squared error: 127.70330782174558\n",
      "\tvariance score: 0.32663361856188255\n",
      "Test stats: \n",
      "\tmean squared error: 159.86910708270102\n",
      "\tvariance score: 0.20360854516971927\n"
     ]
    }
   ],
   "source": [
    "# initialize linear regression object\n",
    "lr = LinearRegression(fit_intercept=True)\n",
    "\n",
    "# fit data to the model\n",
    "lr.fit(x_train, y_train)\n",
    "\n",
    "# make prediction\n",
    "predicted_train = lr.predict(x_train)\n",
    "predicted_test  = lr.predict(x_test)\n",
    "\n",
    "print('Coefficients: \\n', lr.coef_)\n",
    "print('Train stats: ')\n",
    "print(\"\\tmean squared error: {}\".format(mean_squared_error(y_true=y_train, y_pred=predicted_train)))\n",
    "print('\\tvariance score: {}'.format(r2_score(y_true=y_train, y_pred=predicted_train)))\n",
    "\n",
    "print('Test stats: ')\n",
    "print(\"\\tmean squared error: {}\".format(mean_squared_error(y_true=y_test, y_pred=predicted_test)))\n",
    "print('\\tvariance score: {}'.format(r2_score(y_true=y_test, y_pred=predicted_test)))\n",
    "\n",
    "\n",
    "# plot outputs\n",
    "# plt.figure(figsize=(10,5))\n",
    "# plt.scatter(x_test, y_test,  color='blue')\n",
    "# plt.plot(x_test, predicted_test, color='red', linewidth=1)\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the residuals after fitting a linear model\n",
    "plt.scatter(predicted_test, predicted_test-y_test)\n",
    "plt.hlines(y=0, xmin=80, xmax=110)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ridge Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Coefficients: \n",
      " [[ 0.18282166  4.50098872  0.76949952  0.12665478 -0.08716232 -0.16378714\n",
      "  -2.14914904  2.89456272  0.19076269]]\n",
      "Train stats: \n",
      "\tmean squared error: 127.70397800443833\n",
      "\tvariance score: 0.3266300847576107\n",
      "Test stats: \n",
      "\tmean squared error: 159.9841825265005\n",
      "\tvariance score: 0.20303529432860912\n"
     ]
    }
   ],
   "source": [
    "# initialize linear regression object\n",
    "ridge_lr = Ridge(fit_intercept=True, alpha=0.5)\n",
    "\n",
    "# fit data to the model\n",
    "ridge_lr.fit(x_train, y_train)\n",
    "\n",
    "# make prediction\n",
    "predicted_train = ridge_lr.predict(x_train)\n",
    "predicted_test  = ridge_lr.predict(x_test)\n",
    "\n",
    "print('Coefficients: \\n', ridge_lr.coef_)\n",
    "print('Train stats: ')\n",
    "print(\"\\tmean squared error: {}\".format(mean_squared_error(y_true=y_train, y_pred=predicted_train)))\n",
    "print('\\tvariance score: {}'.format(r2_score(y_true=y_train, y_pred=predicted_train)))\n",
    "\n",
    "print('Test stats: ')\n",
    "print(\"\\tmean squared error: {}\".format(mean_squared_error(y_true=y_test, y_pred=predicted_test)))\n",
    "print('\\tvariance score: {}'.format(r2_score(y_true=y_test, y_pred=predicted_test)))\n",
    "\n",
    "\n",
    "# plot outputs\n",
    "# plt.figure(figsize=(10,5))\n",
    "# plt.scatter(x_test, y_test,  color='blue')\n",
    "# plt.plot(x_test, predicted_test, color='red', linewidth=1)\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Lasso Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Coefficients: \n",
      " [0.36553191]\n",
      "Train stats: \n",
      "\tmean squared error: 166.7475510080217\n",
      "\tvariance score: 0.12075734801898497\n",
      "Test stats: \n",
      "\tmean squared error: 192.9979433371658\n",
      "\tvariance score: 0.03857652251709365\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# initialize linear regression object\n",
    "lasso_lr = Lasso(fit_intercept=True, alpha=0.5)\n",
    "\n",
    "# fit data to the model\n",
    "lasso_lr.fit(x_train, y_train)\n",
    "\n",
    "# make prediction\n",
    "predicted_train = lasso_lr.predict(x_train)\n",
    "predicted_test  = lasso_lr.predict(x_test)\n",
    "\n",
    "print('Coefficients: \\n', lasso_lr.coef_)\n",
    "print('Train stats: ')\n",
    "print(\"\\tmean squared error: {}\".format(mean_squared_error(y_true=y_train, y_pred=predicted_train)))\n",
    "print('\\tvariance score: {}'.format(r2_score(y_true=y_train, y_pred=predicted_train)))\n",
    "\n",
    "print('Test stats: ')\n",
    "print(\"\\tmean squared error: {}\".format(mean_squared_error(y_true=y_test, y_pred=predicted_test)))\n",
    "print('\\tvariance score: {}'.format(r2_score(y_true=y_test, y_pred=predicted_test)))\n",
    "\n",
    "\n",
    "# plot outputs\n",
    "plt.figure(figsize=(10,5))\n",
    "plt.scatter(x_test, y_test,  color='blue')\n",
    "plt.plot(x_test, predicted_test, color='red', linewidth=1)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## kNN Regression "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train stats: \n",
      "\tmean squared error: 139.26931471725442\n",
      "\tvariance score: 0.265647256158583\n",
      "Test stats: \n",
      "\tmean squared error: 182.0038309736111\n",
      "\tvariance score: 0.09334393380468897\n"
     ]
    }
   ],
   "source": [
    "knn_r = KNeighborsRegressor(n_neighbors=40)\n",
    "\n",
    "# fit data to the model\n",
    "knn_r.fit(x_train, y_train)\n",
    "\n",
    "# make prediction\n",
    "predicted_train = knn_r.predict(x_train)\n",
    "predicted_test  = knn_r.predict(x_test)\n",
    "\n",
    "print('Train stats: ')\n",
    "print(\"\\tmean squared error: {}\".format(mean_squared_error(y_true=y_train, y_pred=predicted_train)))\n",
    "print('\\tvariance score: {}'.format(r2_score(y_true=y_train, y_pred=predicted_train)))\n",
    "\n",
    "print('Test stats: ')\n",
    "print(\"\\tmean squared error: {}\".format(mean_squared_error(y_true=y_test, y_pred=predicted_test)))\n",
    "print('\\tvariance score: {}'.format(r2_score(y_true=y_test, y_pred=predicted_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Trees"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train stats: \n",
      "\tmean squared error: 0.0\n",
      "\tvariance score: 1.0\n",
      "Test stats: \n",
      "\tmean squared error: 379.5539288888889\n",
      "\tvariance score: -0.8907562013091352\n"
     ]
    }
   ],
   "source": [
    "dt_r = DecisionTreeRegressor()\n",
    "\n",
    "# fit data to the model\n",
    "dt_r.fit(x_train, y_train)\n",
    "\n",
    "# make prediction\n",
    "predicted_train = dt_r.predict(x_train)\n",
    "predicted_test  = dt_r.predict(x_test)\n",
    "\n",
    "print('Train stats: ')\n",
    "print(\"\\tmean squared error: {}\".format(mean_squared_error(y_true=y_train, y_pred=predicted_train)))\n",
    "print('\\tvariance score: {}'.format(r2_score(y_true=y_train, y_pred=predicted_train)))\n",
    "\n",
    "print('Test stats: ')\n",
    "print(\"\\tmean squared error: {}\".format(mean_squared_error(y_true=y_test, y_pred=predicted_test)))\n",
    "print('\\tvariance score: {}'.format(r2_score(y_true=y_test, y_pred=predicted_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train stats: \n",
      "\tmean squared error: 146.40968762486773\n",
      "\tvariance score: 0.22799680568137437\n",
      "Test stats: \n",
      "\tmean squared error: 240.77323205133393\n",
      "\tvariance score: -0.19941712352441932\n"
     ]
    }
   ],
   "source": [
    "rf_r = RandomForestRegressor(n_estimators=100)\n",
    "\n",
    "# fit data to the model\n",
    "rf_r.fit(x_train, y_train.values.reshape(len(y_train),))\n",
    "\n",
    "# make prediction\n",
    "predicted_train = rf_r.predict(x_train)\n",
    "predicted_test  = rf_r.predict(x_test)\n",
    "\n",
    "print('Train stats: ')\n",
    "print(\"\\tmean squared error: {}\".format(mean_squared_error(y_true=y_train, y_pred=predicted_train)))\n",
    "print('\\tvariance score: {}'.format(r2_score(y_true=y_train, y_pred=predicted_train)))\n",
    "\n",
    "print('Test stats: ')\n",
    "print(\"\\tmean squared error: {}\".format(mean_squared_error(y_true=y_test, y_pred=predicted_test)))\n",
    "print('\\tvariance score: {}'.format(r2_score(y_true=y_test, y_pred=predicted_test)))"
   ]
  }
 ],
 "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}