DSI Instructor Challenge submission

Task_Modelling&Debug.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 290,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import csv\n",
    "\n",
    "with open(\"field_names.txt\",\"r\") as f:\n",
    "     field_names= f.read().splitlines()\n",
    "with open('breast-cancer.csv',newline='') as f:\n",
    "    r = csv.reader(f)\n",
    "    data = [line for line in r]\n",
    "with open('breast-cancer.csv','w',newline='') as f:\n",
    "    w = csv.writer(f)\n",
    "    w.writerow(field_names)\n",
    "    w.writerows(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 291,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Index(['ID', 'diagnosis', 'radius_mean', 'radius_sd_error', 'radius_worst',\n",
      "       'texture_mean', 'texture_sd_error', 'texture_worst', 'perimeter_mean',\n",
      "       'perimeter_sd_error', 'perimeter_worst', 'area_mean', 'area_sd_error',\n",
      "       'area_worst', 'smoothness_mean', 'smoothness_sd_error',\n",
      "       'smoothness_worst', 'compactness_mean', 'compactness_sd_error',\n",
      "       'compactness_worst', 'concavity_mean', 'concavity_sd_error',\n",
      "       'concavity_worst', 'concave_points_mean', 'concave_points_sd_error',\n",
      "       'concave_points_worst', 'symmetry_mean', 'symmetry_sd_error',\n",
      "       'symmetry_worst', 'fractal_dimension_mean',\n",
      "       'fractal_dimension_sd_error', 'fractal_dimension_worst'],\n",
      "      dtype='object')\n"
     ]
    }
   ],
   "source": [
    "#load as df\n",
    "data = pd.read_csv('breast-cancer.csv')\n",
    "print(data.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Python Coding and Data Set\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean smoothness for Mallignant cells - 4.3239292452830185\n",
      "Median smoothness for Mallignant cells - 3.6795\n",
      "Mean smoothness for Benign cells - 2.0003212885154062\n",
      "Median smoothnes for Benign cells - 1.851\n"
     ]
    }
   ],
   "source": [
    "# Mean smoothness for Mallignant cells\n",
    "print('Mean smoothness for Mallignant cells -',np.mean(data[data['diagnosis']=='M']['smoothness_mean']))\n",
    "\n",
    "# Median smoothnes for Mallignant cells\n",
    "print('Median smoothness for Mallignant cells -', np.median(data[data['diagnosis']=='M']['smoothness_mean']))\n",
    "\n",
    "# Mean smoothness for Benign cells\n",
    "print('Mean smoothness for Benign cells -',np.mean(data[data['diagnosis']=='B']['smoothness_mean']))\n",
    "\n",
    "# Median smoothnes for Benign cells\n",
    "print('Median smoothnes for Benign cells -',np.median(data[data['diagnosis']=='B']['smoothness_mean']))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean compactness for Mallignant cells - 0.03228116509433963\n",
      "Median compactness for Mallignant cells - 0.02859\n",
      "Mean compactness for Benign cells - 0.021438246498599437\n",
      "Median compactness for Benign cells - 0.016309999999999998\n"
     ]
    }
   ],
   "source": [
    "# Mean compactness for Mallignant cells\n",
    "print('Mean compactness for Mallignant cells -',np.mean(data[data['diagnosis']=='M']['compactness_mean']))\n",
    "\n",
    "# Median compactness for Mallignant cells\n",
    "print('Median compactness for Mallignant cells -', np.median(data[data['diagnosis']=='M']['compactness_mean']))\n",
    "\n",
    "# Mean compactness for Benign cells\n",
    "print('Mean compactness for Benign cells -',np.mean(data[data['diagnosis']=='B']['compactness_mean']))\n",
    "\n",
    "# Median compactness for Benign cells\n",
    "print('Median compactness for Benign cells -',np.median(data[data['diagnosis']=='B']['compactness_mean']))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 294,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Size of sampled data (30, 32)\n"
     ]
    }
   ],
   "source": [
    "# generating bootstrap samples\n",
    "\n",
    "def sample_data(size, dataset):\n",
    "    sample_data = dataset.values[np.random.randint(size, size=size)]\n",
    "    return sample_data\n",
    "\n",
    "#example\n",
    "s_data = sample_data(30,data)\n",
    "\n",
    "print('Size of sampled data',s_data.shape)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exploratory Analysis\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 295,
   "metadata": {},
   "outputs": [],
   "source": [
    "# encoding 'diagnosis' feature as 0/1\n",
    "from sklearn.preprocessing import LabelEncoder \n",
    "le = LabelEncoder()\n",
    "le.fit(data['diagnosis'])\n",
    "data['diagnosis']=le.transform(data['diagnosis'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 307,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "concavity_worst\n",
      "perimeter_sd_error\n",
      "concave_points_sd_error\n",
      "fractal_dimension_mean\n"
     ]
    }
   ],
   "source": [
    "# print top 4 correlation coefficients\n",
    "for i in sorted(corr_numeric[\"diagnosis\"])[-5:]:\n",
    "    if i != 1.0:\n",
    "        print(corr_numeric[corr_numeric[\"diagnosis\"] == i].index.values[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x1ca17407d08>"
      ]
     },
     "execution_count": 308,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 180x720 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "sns.set(style=\"ticks\")\n",
    "sns.pairplot(data, x_vars=['diagnosis'], y_vars=['perimeter_sd_error','concavity_worst','concave_points_sd_error','fractal_dimension_mean'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 312,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1080 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "fig = plt.figure(figsize=(20,15))\n",
    "corr_numeric = data.corr()\n",
    "\n",
    "sns.heatmap(corr_numeric, cbar=True, cmap=\"RdBu_r\")\n",
    "plt.title(\"Correlation Matrix\", fontsize=18)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Modeling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 313,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(569, 31)\n",
      "(569,)\n",
      "(381, 31) (188, 31) (381,) (188,)\n",
      "0    235\n",
      "1    146\n",
      "Name: diagnosis, dtype: int64 0    122\n",
      "1     66\n",
      "Name: diagnosis, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X = data.loc[: , data.columns != 'diagnosis']\n",
    "print(X.shape)\n",
    "\n",
    "Y = data['diagnosis']\n",
    "print(Y.shape)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.33, random_state=5)\n",
    "print(X_train.shape, X_test.shape, y_train.shape, y_test.shape)\n",
    "\n",
    "# checking uniform distribution of y labels\n",
    "print(y_train.value_counts(), y_test.value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 314,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.35106382978723405\n",
      "[[  0 122]\n",
      " [  0  66]]\n",
      "0.38320209973753283\n",
      "[[  0 235]\n",
      " [  0 146]]\n"
     ]
    }
   ],
   "source": [
    "# as we had seen there is strong correlations between predicted variables and target variables. \n",
    "# Logistic Reg won't help us solving this problem\n",
    "# dataset violates LR assumptions\n",
    "\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "#random state to allow reproducibility of results\n",
    "LR_model= LogisticRegression(random_state=0)\n",
    "LR_model.fit(X_train, y_train)\n",
    "\n",
    "y_pred = LR_model.predict(X_test)\n",
    "\n",
    "from sklearn.metrics import *\n",
    "\n",
    "print(accuracy_score(y_test, y_pred))\n",
    "print(confusion_matrix(y_test, y_pred))\n",
    "\n",
    "\n",
    "# checking if model overfits\n",
    "\n",
    "y_pred_train = LR_model.predict(X_train)\n",
    "\n",
    "print(accuracy_score(y_train, y_pred_train))\n",
    "print(confusion_matrix(y_train, y_pred_train))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summary for Logistic Regression\n",
    "\n",
    "1. Should always be first model choice for binary classification. \n",
    "2. Has some assumptions like no strong correlation among predictor variables i.e. multi-collinearity\n",
    "3. Above model completely mis-classifies 0 i.e. benign class, as seen in confusion matrix.\n",
    "\n",
    "4. Possible steps to improve model can be removing multi-collinearity i.e. removing variables having strong correlation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 315,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, ccp_alpha=0.0, class_weight=None,\n",
       "                       criterion='gini', max_depth=4, max_features='auto',\n",
       "                       max_leaf_nodes=None, max_samples=None,\n",
       "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                       min_samples_leaf=1, min_samples_split=2,\n",
       "                       min_weight_fraction_leaf=0.0, n_estimators=100,\n",
       "                       n_jobs=None, oob_score=False, random_state=0, verbose=0,\n",
       "                       warm_start=False)"
      ]
     },
     "execution_count": 315,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "clf = RandomForestClassifier(max_depth=4, random_state=0)\n",
    "clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 316,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train acc - 0.994750656167979\n",
      "Train conf matrix \n",
      " [[235   0]\n",
      " [  2 144]] \n",
      "-----\n",
      "Train acc - 0.973404255319149\n",
      "Train conf matrix \n",
      " [[120   2]\n",
      " [  3  63]] \n",
      "-----\n",
      "Accuracy: 0.96 (+/- 0.02)\n"
     ]
    }
   ],
   "source": [
    "# checking for overfitting (train)\n",
    "y_train_pred = clf.predict(X_train)\n",
    "print('Train acc -',accuracy_score(y_train, y_train_pred))\n",
    "print('Train conf matrix \\n',confusion_matrix(y_train, y_train_pred), '\\n-----')\n",
    "\n",
    "# Testing \n",
    "y_pred = clf.predict(X_test)\n",
    "print('Train acc -', accuracy_score(y_test,y_pred))\n",
    "# test accuracy < train accuracy (good to go)\n",
    "print('Train conf matrix \\n', confusion_matrix(y_test, y_pred), '\\n-----')\n",
    "\n",
    "# Control for overfitting\n",
    "# Cross validation performed for more rigorous analysis (check against overfitting)\n",
    "\n",
    "from sklearn.model_selection import cross_val_score\n",
    "scores = cross_val_score(clf, X, Y, cv=3)\n",
    "print(\"Accuracy: %0.2f (+/- %0.2f)\" % (scores.mean(), scores.std() * 2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 317,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top 3 contributing variables are -- \n",
      "concavity_worst\n",
      "fractal_dimension_mean\n",
      "concave_points_sd_error\n"
     ]
    }
   ],
   "source": [
    "# checking top 3 highest contributing variables\n",
    "\n",
    "ind_top = clf.feature_importances_.argsort()[-3:][::-1]\n",
    "print('Top 3 contributing variables are -- ')\n",
    "for i in ind_top:\n",
    "    print(X_train.columns[i])\n",
    "\n",
    "# Note - this result is line with our EDA above for correaltion diagram\n",
    "# Top contributing varibales are same "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summary for Random Forest\n",
    "\n",
    "1. Random Forest is one of the most widely used ML algorithms. Its a type of ensemble model where multiple weak predictors are bundled to make 1 strong prediction. \n",
    "2. RF is an ensemble of multiple decision trees. It uses averaging to improve prediction accuracy and control overfitting.\n",
    "\n",
    "3. Due to paucity of time & data both, I haven't performed any hyper-parameter tuning. GridSearchCV is a very good way to perform HP optimization here. Jump in accuracy by 0.5 to 1 %.\n",
    "4. Feature importances are a built-in feature in RF that gives you the top contributing factor here."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## GBM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 318,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradientBoostingClassifier(ccp_alpha=0.0, criterion='friedman_mse', init=None,\n",
       "                           learning_rate=0.1, loss='deviance', max_depth=3,\n",
       "                           max_features=None, max_leaf_nodes=None,\n",
       "                           min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                           min_samples_leaf=1, min_samples_split=2,\n",
       "                           min_weight_fraction_leaf=0.0, n_estimators=100,\n",
       "                           n_iter_no_change=None, presort='deprecated',\n",
       "                           random_state=0, subsample=1.0, tol=0.0001,\n",
       "                           validation_fraction=0.1, verbose=0,\n",
       "                           warm_start=False)"
      ]
     },
     "execution_count": 318,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "\n",
    "GBC_clf = GradientBoostingClassifier(random_state=0)\n",
    "GBC_clf.fit(X_train,y_train)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 319,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training acc - 1.0\n",
      "Train conf matrix \n",
      " [[235   0]\n",
      " [  0 146]] \n",
      "-----\n",
      "Test acc - 0.9680851063829787\n",
      "Test conf matrix \n",
      " [[120   2]\n",
      " [  4  62]] \n",
      "-----\n",
      "CV Accuracy: 0.96 (+/- 0.04)\n"
     ]
    }
   ],
   "source": [
    "y_train_pred = GBC_clf.predict(X_train)\n",
    "print('Training acc -',accuracy_score(y_train,y_train_pred))\n",
    "print('Train conf matrix \\n',confusion_matrix(y_train,y_train_pred), '\\n-----')\n",
    "\n",
    "# no false positives/negatives -> indicating overfitting\n",
    "# additionally 1 classificaito score also indicating same\n",
    "\n",
    "y_test_pred = GBC_clf.predict(X_test)\n",
    "print('Test acc -',accuracy_score(y_test, y_test_pred))\n",
    "print('Test conf matrix \\n',confusion_matrix(y_test, y_test_pred), '\\n-----')\n",
    "\n",
    "# Control for overfitting\n",
    "# Cross validation performed for more rigorous analysis (check against overfitting)\n",
    "\n",
    "from sklearn.model_selection import cross_val_score\n",
    "scores = cross_val_score(GBC_clf, X, Y, cv=4)\n",
    "print(\"CV Accuracy: %0.2f (+/- %0.2f)\" % (scores.mean(), scores.std() * 2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 320,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top 3 contributing variables are -- \n",
      "perimeter_sd_error\n",
      "fractal_dimension_mean\n",
      "concave_points_worst\n"
     ]
    }
   ],
   "source": [
    "# checking top 3 highest contributing variables\n",
    "\n",
    "ind_top = GBC_clf.feature_importances_.argsort()[-3:][::-1]\n",
    "print('Top 3 contributing variables are -- ')\n",
    "for i in ind_top:\n",
    "    print(X_train.columns[i])\n",
    "    \n",
    "# pretty much inline with our above analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summary for GBM\n",
    "\n",
    "1. Along with Random Forest, GBM is the most widely used ML algorithm. GBC is also an ensemble type model that uses decision trees. \n",
    "2. But unlike RF which builds the trees alltogether, GBC builds the trees step by step. Hence it is an additive model. \n",
    "\n",
    "3. In our above analysis we see that our model overfits during training i.e. accuracy 100. This is because of less data. Ensemble models require lots of data.\n",
    "4. Feature importance is part of the model itself here."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## --"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Part 2 "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Sample 1.py"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "#from sklearn import LinearRegression\n",
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "#from sklearn.cross_validation import cross_val_score\n",
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "# Load data\n",
    "d = pd.read_csv('train.csv')\n",
    "\n",
    "# Setup data for prediction\n",
    "\n",
    "# incorrect reference of dataframe \"data\". Correct reference is 'd'\n",
    "#x1 = data.SalaryNormalized\n",
    "x1 = d.SalaryNormalized\n",
    "\n",
    "# same issue as above\n",
    "#x2 = pd.get_dummies(data.ContractType)\n",
    "x2 = pd.get_dummies(d.ContractType)\n",
    "\n",
    "## Since this is the predicted varaible, name it y (industry standard)\n",
    "## Corrected version\n",
    "##y = pd.get_dummies(d.ContractType)\n",
    "\n",
    "# Setup model\n",
    "model = LinearRegression()\n",
    "\n",
    "# Evaluate model\n",
    "\n",
    "# Please re-access your model requirements i.e.\n",
    "# what is it that you want to predict i.e. target variable (Y)\n",
    "# identify your inputs i.e. predictor variables (X)\n",
    "# using above variable for your model is a bad design choice. Will give very poor results.\n",
    "\n",
    "# incorrect import statement \n",
    "from sklearn.model_selection import cross_val_score\n",
    "#from sklearn.cross_validation import cross_val_score\n",
    "\n",
    "# incorrect import statement \n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# using cv=1, beats the purpose of k-fold cross validation\n",
    "# when cv=1, it effectively becomes a simple test-train split.\n",
    "# Which doesn't help in reducing overfitting\n",
    "\n",
    "# Some changes in attribute order for cross_val_score\n",
    "# The correct format for cross_val_score is cross_val_score(estimator, X, y=None....)\n",
    "# updating attribute names x1=>x , x2=>Y\n",
    "# 'scoring' parameter doesn't define the type of error metric i.e. incorrect attribute value\n",
    "scores = cross_val_score(model, x2, x1, cv=1, scoring='mean_absolute_error')\n",
    "\n",
    "# you can read more about cross_val_score on sklearn website https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.cross_val_score.html\n",
    "\n",
    "## correct format --\n",
    "## scores = cross_val_score(model, x1, y, cv=5)\n",
    "\n",
    "print(scores.mean())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Sample2.py\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "from sklearn.cross_validation import cross_val_score\n",
    "#from sklearn.model_selection import cross_val_score\n",
    "\n",
    "# Load data\n",
    "data = pd.read_csv('train.csv')\n",
    "\n",
    "\n",
    "# Setup data for prediction\n",
    "\n",
    "y = data.SalaryNormalized\n",
    "X = pd.get_dummies(data.ContractType)\n",
    "\n",
    "# Setup model\n",
    "model = LinearRegression()\n",
    "\n",
    "# Evaluate model\n",
    "\n",
    "# Please re-access your model requirements i.e.\n",
    "# what is it that you want to predict i.e. target variable (Y)\n",
    "# identify your inputs i.e. predictor variables (X)\n",
    "# using above variable for your model is a bad design choice. Will give very poor results.\n",
    "\n",
    "# Some changes in attribute order for cross_val_score\n",
    "# The correct format for cross_val_score is cross_val_score(estimator, X, y=None....)\n",
    "# updating attribute names x1=>x , x2=>Y\n",
    "# 'scoring' parameter doesn't define the type of error metric i.e. incorrect attribute value\n",
    "\n",
    "scores = cross_val_score(model, X, y, cv=5, scoring='mean_absolute_error')\n",
    "\n",
    "\n",
    "# you can read more about cross_val_score on sklearn website https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.cross_val_score.html\n",
    "\n",
    "## correct format --\n",
    "## scores = cross_val_score(model, x1, y, cv=5)\n",
    "print(scores.mean())"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "TF_GPU",
   "language": "python",
   "name": "tf_gpu"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}