Detecting Intentional Misstatements by US Corporations.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Detecting Intentional Misstatements by US Corporations\n",
    "\n",
    "**Motivation and objective:** Predicting whether a US corporation is intentionally misstating its financial statement is a supervised binary classification problem with imbalanced class distribution. This notebook aims to find the best machine learning model and hyperparameters to achieve a higher AUC score. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model Selection & Implementation\n",
    "\n",
    "The content of this notebook is as follow:\n",
    "\n",
    "- Dealing with missing data for test and train data set \n",
    "- Model selection and implementation\n",
    "- Model hyperparameter tuning with Randomized Grid CV search "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#importing the relevant libraries\n",
    "import pandas as pd \n",
    "import numpy as np \n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns \n",
    "%matplotlib inline\n",
    "\n",
    "pd.set_option('display.max_colwidth',-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import testing and train data that contains all variables engineered \n",
    "training = pd.read_csv('training-merged.csv')\n",
    "testing = pd.read_csv('testing-merged.csv')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Class distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Imbalance class distribution: 0.02386117136659436\n"
     ]
    }
   ],
   "source": [
    "per_of_restatement  =len(training[training.Restate_Int == 1])/len(training)\n",
    "print('Imbalance class distribution:', per_of_restatement)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dealing with missing values \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Remove all the rows with missing values in any features before variable selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "training.dropna(axis = 0, inplace= True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Variable Selection \n",
    "\n",
    "Based on the features identified from our exploratory data analytics, here are the variables that we will be including in our model: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# variable selection here\n",
    "#creat X_train, y_train, X_test \n",
    "X_train = training[['Coleman_Liau_Index', 'averageWordsPerSentence', 'FinTerms_Negative',\n",
    "       'FinTerms_Positive', 'FinTerms_Litigious', 'FinTerms_ModalStrong',\n",
    "       'log_total_assets', 'change_in_inventory', 'percent_soft_assets',\n",
    "       'merger', 'big_n_auditor', 'tot_financing', 'ex_ante_financing',\n",
    "       'restructuring', 'auop_qualified', 'auopic_no_report',\n",
    "       'auopic_effective', 'auopic_adverse', 'au_9', 'ch_au', 'sic_2092',\n",
    "       'sic_2253', 'sic_3524', 'sic_3569', 'sic_4013', 'sic_4512', 'sic_5122',\n",
    "       'sic_5399', 'sic_6552', 'sic_8093','GOING_CONCERN']].copy()\n",
    "\n",
    "y_train= training['Restate_Int'].copy()\n",
    "\n",
    "X_test = testing[['Coleman_Liau_Index', 'averageWordsPerSentence', 'FinTerms_Negative',\n",
    "       'FinTerms_Positive', 'FinTerms_Litigious', 'FinTerms_ModalStrong',\n",
    "       'log_total_assets', 'change_in_inventory', 'percent_soft_assets',\n",
    "       'merger', 'big_n_auditor', 'tot_financing', 'ex_ante_financing',\n",
    "       'restructuring', 'auop_qualified', 'auopic_no_report',\n",
    "       'auopic_effective', 'auopic_adverse', 'au_9', 'ch_au', 'sic_2092',\n",
    "       'sic_2253', 'sic_3524', 'sic_3569', 'sic_4013', 'sic_4512', 'sic_5122',\n",
    "       'sic_5399', 'sic_6552', 'sic_8093','GOING_CONCERN']]\n",
    "\n",
    "output_gvkey = testing['Restate_Int']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Replace all testing feature missing values: \n",
    "- if the feature is categorical, replace with most frequent value\n",
    "- if the feature is numeric, replace with mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# fill in all the missing values \n",
    "from sklearn.base import TransformerMixin\n",
    "class DataFrameImputer(TransformerMixin):\n",
    "    def __init__(self):\n",
    "        \"\"\"Impute missing values.\n",
    "        Columns of dtype object are imputed with the most frequent value \n",
    "        in column.\n",
    "        Columns of other types are imputed with mean of column.\n",
    "        \"\"\"\n",
    "    def fit(self, X, y=None):\n",
    "        self.fill = pd.Series([X[c].value_counts().index[0]\n",
    "            if X[c].dtype == np.dtype('O') else X[c].mean() for c in X],\n",
    "            index=X.columns)\n",
    "        return self\n",
    "    def transform(self, X, y=None):\n",
    "        return X.fillna(self.fill)\n",
    "X = pd.DataFrame(X_test)\n",
    "X_test = DataFrameImputer().fit_transform(X)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14287\n",
      "14287\n",
      "2989\n",
      "2989\n"
     ]
    }
   ],
   "source": [
    "#make sure the length of the row matches\n",
    "print(len(X_train))\n",
    "print(len(y_train))\n",
    "print(len(X_test))\n",
    "print(len(output_gvkey))\n",
    "\n",
    "X_train.to_csv('X_train.csv')\n",
    "X_test.to_csv('X_test.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Implementation of Model "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Algorithm 1:  Random forest"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Rational for selecting random forest: ** \n",
    "- Can be used for both continuos or categorical target prediction problem\n",
    "- Adopts bagging technique to combine multiples decision trees formed from subsets of the training set to improve the overall performance of the model \n",
    "- Bagging technique ensures minimum correlation among the trees which prevents overfitting, as tree methods used on its own tends to overfit\n",
    "- Good at detecting interactions between different features, but highly correlated features can mask these interactions\n",
    "- Able to handle a high number of features \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prediction with Random Forest's Default Setting\n",
    "With the default hyperparameters in Random Forest, the in sample accuracy is about 0.58 with 5 fold CV\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import KFold\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "import xgboost as xgb\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Default hyperparameters in Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'bootstrap': True,\n",
       " 'class_weight': None,\n",
       " 'criterion': 'gini',\n",
       " 'max_depth': None,\n",
       " 'max_features': 'auto',\n",
       " 'max_leaf_nodes': None,\n",
       " 'min_impurity_decrease': 0.0,\n",
       " 'min_impurity_split': None,\n",
       " 'min_samples_leaf': 1,\n",
       " 'min_samples_split': 2,\n",
       " 'min_weight_fraction_leaf': 0.0,\n",
       " 'n_estimators': 10,\n",
       " 'n_jobs': 1,\n",
       " 'oob_score': False,\n",
       " 'random_state': None,\n",
       " 'verbose': 0,\n",
       " 'warm_start': False}"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# create a random forest object \n",
    "clf=RandomForestClassifier()\n",
    "# check the default paramater of random forest \n",
    "clf.get_params()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average 5-Fold CV Score: 0.5802198327632977\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', 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=10, n_jobs=1,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf=RandomForestClassifier( )\n",
    "\n",
    "cv_scores = cross_val_score(clf, X_train, y_train, scoring = 'roc_auc', cv =5)\n",
    "print(\"Average 5-Fold CV Score: {}\".format(np.mean(cv_scores)))\n",
    "#y_pred=clf.predict(X_test)\n",
    "\n",
    "clf.fit(X_train,y_train)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##  Hyperparameter Tuning in Random Forest with Random Search Cross Validation\n",
    "**Main hyperparameters to find tune:**\n",
    "- n_estimators = number of trees in the foreset\n",
    "- max_features = max number of features considered for splitting a node\n",
    "- max_depth = max number of levels in each decision tree\n",
    "- min_samples_split = min number of data points placed in a node before the node is split\n",
    "- min_samples_leaf = min number of data points allowed in a leaf node\n",
    "- bootstrap = method for sampling data points (with or without replacement)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'n_estimators': [200, 288, 377, 466, 555, 644, 733, 822, 911, 1000], 'max_features': ['auto', 'sqrt'], 'max_depth': [10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, None], 'min_samples_split': [2, 5, 10], 'min_samples_leaf': [1, 2, 4], 'bootstrap': [True, False]}\n"
     ]
    }
   ],
   "source": [
    "# prepare the hyperparameters to tune \n",
    "# Number of trees in random forest\n",
    "n_estimators = [int(x) for x in np.linspace(start = 200, stop = 1000, num = 10)]\n",
    "# Number of features to consider at every split\n",
    "max_features = ['auto', 'sqrt']\n",
    "# Maximum number of levels in tree\n",
    "max_depth = [int(x) for x in np.linspace(10, 110, num = 11)]\n",
    "max_depth.append(None)\n",
    "# Minimum number of samples required to split a node\n",
    "min_samples_split = [2, 5, 10]\n",
    "# Minimum number of samples required at each leaf node\n",
    "min_samples_leaf = [1, 2, 4]\n",
    "# Method of selecting samples for training each tree\n",
    "bootstrap = [True, False]\n",
    "# Create the random grid\n",
    "random_grid = {'n_estimators': n_estimators,\n",
    "               'max_features': max_features,\n",
    "               'max_depth': max_depth,\n",
    "               'min_samples_split': min_samples_split,\n",
    "               'min_samples_leaf': min_samples_leaf,\n",
    "               'bootstrap': bootstrap}\n",
    "\n",
    "print(random_grid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 2 folds for each of 100 candidates, totalling 200 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done  33 tasks      | elapsed:  3.9min\n",
      "[Parallel(n_jobs=-1)]: Done 154 tasks      | elapsed: 18.7min\n",
      "[Parallel(n_jobs=-1)]: Done 200 out of 200 | elapsed: 23.1min finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "RandomizedSearchCV(cv=2, error_score='raise',\n",
       "          estimator=RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', 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=10, n_jobs=1,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False),\n",
       "          fit_params=None, iid=True, n_iter=100, n_jobs=-1,\n",
       "          param_distributions={'n_estimators': [200, 288, 377, 466, 555, 644, 733, 822, 911, 1000], 'max_features': ['auto', 'sqrt'], 'max_depth': [10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, None], 'min_samples_split': [2, 5, 10], 'min_samples_leaf': [1, 2, 4], 'bootstrap': [True, False]},\n",
       "          pre_dispatch='2*n_jobs', random_state=42, refit=True,\n",
       "          return_train_score='warn', scoring='roc_auc', verbose=2)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf=RandomForestClassifier( )\n",
    "rf_random = RandomizedSearchCV(estimator = clf,\n",
    "                               param_distributions = random_grid,\n",
    "                               n_iter = 100,\n",
    "                               cv = 2,\n",
    "                               verbose=2,\n",
    "                               random_state=42,\n",
    "                               n_jobs = -1,\n",
    "                               scoring = 'roc_auc')\n",
    "# Fit the random search model\n",
    "rf_random.fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'n_estimators': 822,\n",
       " 'min_samples_split': 2,\n",
       " 'min_samples_leaf': 4,\n",
       " 'max_features': 'sqrt',\n",
       " 'max_depth': 10,\n",
       " 'bootstrap': True}"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get the best parameter \n",
    "rf_random.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'n_estimators': 822,\n",
       " 'min_samples_split': 2,\n",
       " 'min_samples_leaf': 4,\n",
       " 'max_features': 'sqrt',\n",
       " 'max_depth': 10,\n",
       " 'bootstrap': True}"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_random.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Fit & Predict \n",
    "clf=RandomForestClassifier(n_estimators=822,\n",
    "                           min_samples_split= 2,\n",
    "                           min_samples_leaf=4,\n",
    "                           max_features= 'sqrt',\n",
    "                           max_depth= 10,\n",
    "                           bootstrap= True )\n",
    "clf.fit(X_train,y_train)\n",
    "pred = clf.predict_proba(X_test) \n",
    "#output the result to CSV \n",
    "\n",
    "results_df = pd.DataFrame(data={'gvkey':testing['gvkey'], 'Restate_Int':pred[:,1].tolist()})\n",
    "results_df.to_csv('rf_py.csv',index =False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x22385a282b0>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "feature_imp = pd.Series(clf.feature_importances_,index =X_train.columns).sort_values(ascending=False)\n",
    "feature_imp\n",
    "\n",
    "plt.figure(figsize=(20,10))\n",
    "sns.barplot(x=feature_imp, y=feature_imp.index)\n",
    "# Add labels to your graph\n",
    "plt.rcParams.update({'font.size': 17})\n",
    "plt.rcParams['figure.figsize'] = [23, 20]\n",
    "\n",
    "plt.xlabel('Feature Importance Score')\n",
    "plt.ylabel('Features')\n",
    "plt.title(\"Visualizing Important Features\")\n",
    "plt.legend()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Algorithm 2:  Support Vector Machines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.svm import SVC"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prediction with SVC's Default Setting\n",
    "With the default hyperparameters in SCV, the in sample accuracy is about 0.48 with 5 fold CV"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Rational for selecting random forest: ** \n",
    "- One of the most robust binary classifcaiont models that is able to take in a large number of variales with high sparsity\n",
    "- Robust at dealing with datasets with imbalanced class frequencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'C': 1.0,\n",
       " 'cache_size': 200,\n",
       " 'class_weight': None,\n",
       " 'coef0': 0.0,\n",
       " 'decision_function_shape': 'ovr',\n",
       " 'degree': 3,\n",
       " 'gamma': 'auto',\n",
       " 'kernel': 'rbf',\n",
       " 'max_iter': -1,\n",
       " 'probability': False,\n",
       " 'random_state': None,\n",
       " 'shrinking': True,\n",
       " 'tol': 0.001,\n",
       " 'verbose': False}"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# create a random forest object \n",
    "clf = SVC()\n",
    "# check the default paramater of random forest \n",
    "clf.get_params()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average 5-Fold CV Score: 0.4800514951779018\n"
     ]
    }
   ],
   "source": [
    "clf = SVC(kernel = 'rbf') \n",
    "cv_results = cross_val_score(clf , X_train, y_train, cv =5, scoring = 'roc_auc')\n",
    "print(\"Average 5-Fold CV Score: {}\".format(np.mean(cv_results)))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##  Hyperparameter Tuning in SVC with Random Search Cross Validation\n",
    "\n",
    "\n",
    "**The hyperparameter to fine tune in SVC are:** \n",
    "- Type of kernel \n",
    "- gamme = for non linear hyperplane\n",
    "- C = penality parameter of the error term, tradeoff bet smooth decision boundary and classifying the traning points correctly\n",
    "- Degree= degree of the polynomial used to find the hyperplane to split the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'gamma': [0.1, 1, 10, 100], 'C': [0.1, 1, 10, 100], 'degree': [0, 1, 2, 3, 4, 5, 6]}\n"
     ]
    }
   ],
   "source": [
    "# create the param for grid search \n",
    "gamma = [0.1,1,10,100 ]\n",
    "C = [0.1,1,10,100]\n",
    "degree = [0,1,2,3,4,5,6] #when degree is 1 , the kernel is same as linear \n",
    "\n",
    "param_grid = dict(gamma=gamma, C=C, degree= degree)\n",
    "print(param_grid)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 2 folds for each of 50 candidates, totalling 100 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done  33 tasks      | elapsed:  2.2min\n",
      "[Parallel(n_jobs=-1)]: Done 100 out of 100 | elapsed:  7.4min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best: 0.599750 using {'gamma': 10, 'degree': 1, 'C': 100}\n",
      "Execution time: 484.8979494571686 ms\n"
     ]
    }
   ],
   "source": [
    "import time \n",
    "\n",
    "# initiate a SVC obj\n",
    "clf = SVC() \n",
    "# Grid search \n",
    "grid = RandomizedSearchCV(estimator=clf, \n",
    "                          param_distributions = param_grid,\n",
    "                          n_iter = 50,\n",
    "                          cv = 2,\n",
    "                          verbose=2,\n",
    "                          random_state=42,\n",
    "                          n_jobs = -1,\n",
    "                          scoring = 'roc_auc') #no job to run in parallel, -1 means using all processor\n",
    "\n",
    "start_time = time.time()\n",
    "grid_result = grid.fit(X_train, y_train)\n",
    "\n",
    "print(\"Best: %f using %s\" % (grid_result.best_score_, grid_result.best_params_))\n",
    "print(\"Execution time: \" + str((time.time() - start_time)) + ' ms')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The best parameter selected is gamma of 10, degree of 1 and C of 100. These settings allows SVC to be non linear at fitting the training data set, penalizes error term by 100 which may lead to overfitting and uses 1 degree of polynomial used to find the hyperplane to split the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LibSVM][LibSVM][LibSVM][LibSVM][LibSVM][LibSVM]Average 5-Fold CV Score: 0.5852981260419671\n"
     ]
    }
   ],
   "source": [
    "#kfold = KFold(n_splits=3, random_state=7)\n",
    "\n",
    "clf = SVC( gamma =10 ,  C= 100 ,probability =True, degree =1 , verbose = True) \n",
    "clf.fit(X_train, y_train)\n",
    "\n",
    "\n",
    "cv_results = cross_val_score(clf , X_train, y_train, cv =5, scoring = 'roc_auc')\n",
    "\n",
    "print(\"Average 5-Fold CV Score: {}\".format(np.mean(cv_results)))\n",
    "\n",
    "output_test = clf.predict_proba(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.01019706, 0.02365011, 0.02375276, ..., 0.0244432 , 0.02432189,\n",
       "       0.02442388])"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "output_test[:,1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "# output the file \n",
    "results_df = pd.DataFrame(data={'gvkey':testing['gvkey'], 'Restate_Int':output_test[:,1]})\n",
    "results_df.to_csv('SVM_funetune_prob.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "References: \n",
    "- https://towardsdatascience.com/hyperparameter-tuning-the-random-forest-in-python-using-scikit-learn-28d2aa77dd74\n",
    "- https://medium.com/all-things-ai/in-depth-parameter-tuning-for-svc-758215394769\n",
    "- https://www.datacamp.com/community/tutorials/random-forests-classifier-python\n",
    "- https://statistics.berkeley.edu/sites/default/files/tech-reports/666.pdf"
   ]
  },
  {
   "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}