SVM_Fit_CV.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Support vector machine classifier fitted by cross validation\n",
    "\n",
    " 1. Load clean data into memory, normalize all of variables with z-score method\n",
    " 2. The 4 parameters found through random grid search and then error estimated through cross validation \n",
    "     \n",
    "     - C: Regularization parameter. (l2-regularization)\n",
    "     - gamma: Kernel coefficient for ‘rbf’, ‘poly’ and ‘sigmoid’.\n",
    "     - kernel: Specifies the kernel type to be used in the algorithm\n",
    "     \n",
    " 3. Measure accuracy over test set and AUC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 200 candidates, totalling 1000 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 1000 out of 1000 | elapsed:  1.1min finished\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import os\n",
    "\n",
    "os.chdir('/Users/bwilliams/GoogleDrive/UniversityOfHelsinki/Fall2020/IML/TeamProject/TermProjectIML/bin')\n",
    "from SplitData import *\n",
    "\n",
    "\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.model_selection import cross_val_score, GridSearchCV, KFold, RandomizedSearchCV, train_test_split\n",
    "\n",
    "from sklearn.utils.fixes import loguniform\n",
    "\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import plot_confusion_matrix\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "params = {'C': loguniform(1e0, 1e3),\n",
    " 'gamma': loguniform(1e-4, 1e-3),\n",
    " 'kernel': ['linear','sigmoid','rbf']}\n",
    "\n",
    "\n",
    "\n",
    "svclassifier = SVC(probability=True)\n",
    "\n",
    "search = RandomizedSearchCV(svclassifier, param_distributions=params, random_state=42, n_iter=200, cv=5, verbose=1, n_jobs=1, return_train_score=True)\n",
    "\n",
    "search.fit(X_trainval_norm, y_trainval_bin)\n",
    "svm_params = search.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'C': 11.091998785656282, 'gamma': 0.0008642674399024057, 'kernel': 'linear'}\n",
      "Accuracy on Train: 0.9825581395348837\n",
      "5-Fold CV Accuracy:  0.8982523444160273\n",
      "Accuracy on Test: 0.8372093023255814\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Predicted', ylabel='Actual'>"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "print(search.best_params_)\n",
    "\n",
    "svm_model = search.best_estimator_\n",
    "\n",
    "y_pred_train = svm_model.predict(X_trainval_norm)\n",
    "\n",
    "print(\"Accuracy on Train:\",metrics.accuracy_score(y_trainval_bin, y_pred_train))\n",
    "\n",
    "y_pred = svm_model.predict(X_test_norm)\n",
    "# probabilities of class event\n",
    "y_pred_proba = svm_model.predict_proba(X_test_norm)[:,0]\n",
    "\n",
    "print('5-Fold CV Accuracy: ', search.best_score_)\n",
    "\n",
    "print(\"Accuracy on Test:\",metrics.accuracy_score(y_test_bin, y_pred))\n",
    "\n",
    "\n",
    "import seaborn as sns\n",
    "confusion_matrix = pd.crosstab(y_test_bin, y_pred, rownames=['Actual'], colnames=['Predicted'])\n",
    "sns.heatmap(confusion_matrix, annot=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "y_test_01 = pd.get_dummies(y_test_bin)['event']\n",
    "\n",
    "fpr, tpr, thresholds = metrics.roc_curve(y_test_01, y_pred_proba)\n",
    "auc = metrics.auc(fpr, tpr)\n",
    "\n",
    "plt.plot(fpr,tpr , marker='.', label='DT')\n",
    "plt.title('AUC: '+ str(auc) )\n",
    "plt.plot(fpr, fpr, linestyle='--', label='Baseline')\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Multi class classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 200 candidates, totalling 1000 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'C': 109.53031576544694, 'gamma': 0.0005494254346819604, 'kernel': 'rbf'}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 1000 out of 1000 | elapsed:  3.2min finished\n"
     ]
    }
   ],
   "source": [
    "########################################################\n",
    "# Multiclass\n",
    "\n",
    "params = {'C': loguniform(1e0, 1e3),\n",
    " 'gamma': loguniform(1e-4, 1e-3),\n",
    " 'kernel': ['linear','sigmoid','rbf']}\n",
    "\n",
    "\n",
    "\n",
    "svclassifier = SVC(probability=True)\n",
    "\n",
    "search = RandomizedSearchCV(svclassifier, param_distributions=params, random_state=42, n_iter=200, cv=5, verbose=1, n_jobs=1, return_train_score=True)\n",
    "\n",
    "search.fit(X_trainval_norm, y_trainval)\n",
    "svm_params = search.best_params_\n",
    "\n",
    "print(search.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on Train: 0.8313953488372093\n",
      "5-Fold CV Accuracy:  0.6918158567774937\n",
      "Accuracy on Test: 0.686046511627907\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='Predicted', ylabel='Actual'>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "svm_model = search.best_estimator_\n",
    "\n",
    "y_pred_train = svm_model.predict(X_trainval_norm)\n",
    "\n",
    "print(\"Accuracy on Train:\",metrics.accuracy_score(y_trainval, y_pred_train))\n",
    "\n",
    "y_pred = svm_model.predict(X_test_norm)\n",
    "# probabilities of class event\n",
    "y_pred_proba = svm_model.predict_proba(X_test_norm)[:,0]\n",
    "\n",
    "print('5-Fold CV Accuracy: ', search.best_score_)\n",
    "\n",
    "print(\"Accuracy on Test:\",metrics.accuracy_score(y_test, y_pred))\n",
    "\n",
    "\n",
    "confusion_matrix = pd.crosstab(y_test, y_pred, rownames=['Actual'], colnames=['Predicted'])\n",
    "sns.heatmap(confusion_matrix, annot=True)\n",
    "\n"
   ]
  },
  {
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}