Unbalanced_classification_boost.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Pattern Recognition - IUST - Fall 98\n",
    "## Taha Samavati - 98722134\n",
    "### Second Project - Ensembles"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# imports\n",
    "import numpy as np\n",
    "from sklearn.tree import DecisionTreeClassifier as DT\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.metrics import precision_recall_fscore_support,roc_curve,auc\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "import matplotlib.pyplot as plt\n",
    "from collections import Counter\n",
    "from sklearn.neighbors import NearestNeighbors\n",
    "from sklearn.preprocessing import normalize\n",
    "import math\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.svm import LinearSVC\n",
    "from math import log\n",
    "import random\n",
    "import os\n",
    "import pandas as pd\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Read dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# read data set - vehicle0.dat consisting of 846 data points each with 18 features and a binary label\n",
    "features = np.zeros((846,18))\n",
    "labels = np.zeros((846,),dtype=np.uint8)\n",
    "\n",
    "with open('./vehicle0.dat','r') as file:\n",
    "    a = file.readlines()\n",
    "    for i in range(len(a)):\n",
    "        a[i] = a[i].replace(\" \",\"\")\n",
    "        b = a[i].split(',')\n",
    "        if b[18] == 'positive\\n':\n",
    "            b[18] = 1\n",
    "        elif b[18] == 'negative\\n':\n",
    "            b[18] = 0\n",
    "        features[i,:] = np.array(b[:18])\n",
    "        labels[i] = b[18]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1.AdaBoost M2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "class AdaboostM2():\n",
    "    \n",
    "    def __init__(self, instances, labels, n_clf):\n",
    "        self.X = instances\n",
    "        self.y = labels.tolist()\n",
    "        \n",
    "        self.clf = []\n",
    "        self.n_clf = n_clf\n",
    "            \n",
    "        self.weights = []\n",
    "        self.init_weight= 1.0 / len(self.X)\n",
    "        \n",
    "        for i in range(len(self.X)):# Assign initial weights to datapoints\n",
    "            self.weights.append(self.init_weight)\n",
    "\n",
    "        self.beta = []\n",
    "        \n",
    "    def fit(self):\n",
    "        \n",
    "        for k in range(self.n_clf):\n",
    "            \n",
    "            base_classifier = DT(max_depth=2)# construct base classifier\n",
    "            base_classifier.fit(self.X, self.y, self.weights)\n",
    "\n",
    "            self.clf.append(base_classifier)\n",
    "            \n",
    "            loss = 0\n",
    "            for x,y,w in zip(self.X, self.y, self.weights):\n",
    "                x = x.reshape(1,-1)\n",
    "                probs = self.clf[k].predict_proba(x)[0]\n",
    "                pred = self.clf[k].predict(x)[0]\n",
    "                if y != pred:\n",
    "                    loss+= w * (1 - probs[y] + probs[pred])\n",
    "            loss *= 0.5\n",
    "            beta = loss/(1-loss)\n",
    "            if math.isnan(beta) or beta == 0:\n",
    "                beta = 1\n",
    "            self.beta.append(beta)\n",
    "            \n",
    "            if loss > 0: # update weights if there is a loss\n",
    "                for i in range(len(self.weights)):\n",
    "                    x = self.X[i]\n",
    "                    x = x.reshape(1,-1)\n",
    "                    probs = self.clf[k].predict_proba(x)[0]\n",
    "                    pred = self.clf[k].predict(x)[0]\n",
    "                    #if self.y[i] != pred:\n",
    "                    self.weights[i] = self.weights[i] * (np.power(beta,(0.5*(1 + probs[y] - probs[pred]))))\n",
    "            \n",
    "            # use total sum to normalize sum of weights to 1.0\n",
    "            sum_weight = sum(self.weights)\n",
    "            self.weights = [w / sum_weight for w in self.weights]\n",
    "            \n",
    "    def predict(self, instance):# compute positive class and negative class score\n",
    "        \n",
    "        p_score, n_score = 0 , 0\n",
    "        instance = instance.reshape(1,-1)\n",
    "        \n",
    "        for i in range(self.n_clf):\n",
    "            pred = self.clf[i].predict(instance)[0]\n",
    "            probs = self.clf[i].predict_proba(instance)[0]\n",
    "            if pred == 1:\n",
    "                p_score += np.log(1/self.beta[i])* probs[1]\n",
    "            else:\n",
    "                n_score += np.log(1/self.beta[i])* probs[0]\n",
    "\n",
    "        if n_score <= p_score:\n",
    "            return 1\n",
    "        else:\n",
    "            return 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Precision and recall for ensemble size of : 10\n",
      "ADABOOST: Avg Precision= 0.76 Avg Recall= 0.9\n",
      "=========================================================\n",
      "Precision and recall for ensemble size of : 50\n",
      "ADABOOST: Avg Precision= 0.77 Avg Recall= 0.9\n",
      "=========================================================\n",
      "Precision and recall for ensemble size of : 100\n",
      "ADABOOST: Avg Precision= 0.77 Avg Recall= 0.9\n",
      "=========================================================\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "kf = KFold(n_splits=5,random_state=76,shuffle=True) # 5fold cross validation\n",
    "\n",
    "def train_adaboost(xtrain, ytrain, xtest, ytest, indx, ensemble_size=10, plot_roc=False):\n",
    "\n",
    "    ada = AdaboostM2(features,labels,ensemble_size)\n",
    "    ada.fit()\n",
    "    \n",
    "    preds=[]\n",
    "    for i in range(xtest.shape[0]):\n",
    "        preds.append(ada.predict(xtest[i]))\n",
    "    if plot_roc:\n",
    "        # plot roc curve\n",
    "        fpr, tpr, _ = roc_curve(ytest, preds)\n",
    "        roc_auc = auc(fpr, tpr)\n",
    "\n",
    "        ax[indx].plot(fpr,tpr,label=\"Ensemble Size=\"+str(esize)+\", auc=\"+str(roc_auc))\n",
    "        ax[indx].legend(loc=10)\n",
    "        if indx==5:\n",
    "            plt.show()\n",
    "    \n",
    "    prec,recall,_,_ = precision_recall_fscore_support(preds, ytest)\n",
    "    return prec,recall\n",
    "\n",
    "# For each ensemble size train clf with adaboost algorithm\n",
    "ensemble_size=[10,50,100]\n",
    "avg_prec_ada, avg_rcall_ada = [], []\n",
    "\n",
    "for esize in ensemble_size:\n",
    "    p_ada,r_ada=[],[] # lists for storing precisions and recalls\n",
    "    indx=0\n",
    "    fig, ax = plt.subplots(1,5,figsize=(14,3))\n",
    "    \n",
    "    for train_index, test_index in kf.split(features):\n",
    "        X_train, X_test = features[train_index], features[test_index]\n",
    "        y_train, y_test = labels[train_index], labels[test_index]\n",
    "        # ADABOOST\n",
    "        precision,recall = train_adaboost(X_train, y_train,X_test,y_test, indx, esize)\n",
    "        p_ada.append(precision)\n",
    "        r_ada.append(recall)\n",
    "        indx+=1\n",
    "\n",
    "    print(\"Precision and recall for ensemble size of : \"+str(esize))\n",
    "    avg_prec_ada.append(np.mean(p_ada))\n",
    "    avg_rcall_ada.append(np.mean(r_ada))\n",
    "    print(\"ADABOOST: Avg Precision= \"+str(round(avg_prec_ada[-1],2))+\" Avg Recall= \"+str(round(avg_rcall_ada[-1],2)))\n",
    "    print(\"=========================================================\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. RUSBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "class RUSBoost:\n",
    "    \n",
    "    def __init__(self, instances, labels, n_clf, desired_balance):\n",
    "        \n",
    "        self.X = instances\n",
    "        self.y = labels.tolist()\n",
    "        \n",
    "        self.clf = []\n",
    "        self.n_clf = n_clf\n",
    "            \n",
    "        self.weights = []\n",
    "        self.init_weight= 1.0 / len(self.X)\n",
    "        for i in range(len(self.X)):\n",
    "            self.weights.append(self.init_weight)\n",
    "            \n",
    "        self.ratio = desired_balance\n",
    "        self.beta = []\n",
    "\n",
    "    \n",
    "    def undersampling(self):\n",
    "\n",
    "        # Check the major class\n",
    "\n",
    "        diff = self.y.count(1) > self.y.count(0)\n",
    "        delete_list = []\n",
    "        keep_list =[]\n",
    "        \n",
    "        if  diff: # if positive class is the major class \n",
    "            for i in range(len(self.y)):\n",
    "                # store positive samples in delete data\n",
    "                if self.y[i] == 1:\n",
    "                    delete_data = [i, self.X[i], 1]\n",
    "                    delete_list.append(delete_data)\n",
    "                else:\n",
    "                    keep_data = [i, self.X[i], 0]\n",
    "                    keep_list.append(keep_data)\n",
    "        else:# if negative class is the major class \n",
    "            for i in range(len(self.y)):\n",
    "                # store positive samples in delete data\n",
    "                if self.y[i] == 0:\n",
    "                    delete_data = [i, self.X[i], 0]\n",
    "                    delete_list.append(delete_data)\n",
    "                else:\n",
    "                    keep_data = [i, self.X[i], 1]\n",
    "                    keep_list.append(keep_data)\n",
    "        \n",
    "        # randomly delete major class samples until the desired balance is reached\n",
    "        while len(delete_list) > self.ratio*(len(delete_list)+len(keep_list)):\n",
    "            k = random.choice(range(len(delete_list)))\n",
    "            delete_list.pop(k)\n",
    "        \n",
    "        all_list = delete_list + keep_list\n",
    "        return sorted(all_list, key=lambda x:x[2])\n",
    "    \n",
    "    def predict(self, instance):\n",
    "        \n",
    "        p_score, n_score = 0 , 0\n",
    "        instance = instance.reshape(1,-1)\n",
    "        \n",
    "        for i in range(self.n_clf):\n",
    "            pred = self.clf[i].predict(instance)[0]\n",
    "            probs = self.clf[i].predict_proba(instance)[0]\n",
    "            if pred == 1:\n",
    "                p_score += np.log(1/self.beta[i]) * probs[1]\n",
    "            else:\n",
    "                n_score += np.log(1/self.beta[i]) * probs[0]\n",
    "        if n_score <= p_score:\n",
    "            return 1\n",
    "        else:\n",
    "            return 0\n",
    "        \n",
    "    def fit(self):\n",
    "        for k in range(self.n_clf):\n",
    "\n",
    "            sampled = self.undersampling()\n",
    "\n",
    "            sampled_weight = [self.weights[s[0]] for s in sampled]\n",
    "            sampled_X      = [s[1] for s in sampled]\n",
    "            sampled_Y       = [s[2] for s in sampled]\n",
    "            \n",
    "            base_classifier = DT(max_depth=2)\n",
    "            base_classifier.fit(sampled_X, sampled_Y, sampled_weight)\n",
    "            \n",
    "            self.clf.append(base_classifier)\n",
    "\n",
    "            loss = 0\n",
    "            for x,y,w in zip(self.X, self.y, self.weights):\n",
    "                x = x.reshape(1,-1)\n",
    "                probs = self.clf[k].predict_proba(x)[0]\n",
    "                pred = self.clf[k].predict(x)[0]\n",
    "                if y != pred:\n",
    "                    loss+= w * (1 - probs[y] + probs[pred])\n",
    "            loss *= 0.5\n",
    "            beta = loss/(1-loss)\n",
    "            if math.isnan(beta) or beta == 0:\n",
    "                beta = 1\n",
    "            self.beta.append(beta)\n",
    "            if loss > 0: # update weights if there is a loss\n",
    "                \n",
    "                for i in range(len(self.weights)):\n",
    "                    x = self.X[i]\n",
    "                    x = x.reshape(1,-1)\n",
    "                    probs = self.clf[k].predict_proba(x)[0]\n",
    "                    pred = self.clf[k].predict(x)[0]\n",
    "                    #if self.y[i] != pred:\n",
    "                    self.weights[i] = self.weights[i] * (np.power(beta,(0.5*(1 + probs[y] - probs[pred]))))\n",
    "            \n",
    "            # use total sum to normalize sum of weights to 1.0\n",
    "            sum_weight = sum(self.weights)\n",
    "            self.weights = [w / sum_weight for w in self.weights]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Precision and recall for ensemble size of : 10\n",
      "RUSBOOST: Avg Precision= 0.82 Avg Recall= 0.75\n",
      "=========================================================\n",
      "Precision and recall for ensemble size of : 50\n",
      "RUSBOOST: Avg Precision= 0.81 Avg Recall= 0.76\n",
      "=========================================================\n",
      "Precision and recall for ensemble size of : 100\n",
      "RUSBOOST: Avg Precision= 0.81 Avg Recall= 0.75\n",
      "=========================================================\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def train_rusboost(xtrain, ytrain, xtest, ytest, indx, ensemble_size=10, plot_roc=False):\n",
    "    \n",
    "    rate = 0.5\n",
    "    rus = RUSBoost(xtrain, ytrain, ensemble_size, rate)\n",
    "    rus.fit()\n",
    "    \n",
    "    preds=[]\n",
    "    for i in range(xtest.shape[0]):\n",
    "        preds.append(rus.predict(xtest[i]))\n",
    "    if plot_roc:\n",
    "        # plot ROC curves\n",
    "        fpr, tpr, _ = roc_curve(ytest, preds)\n",
    "        roc_auc = auc(fpr, tpr)\n",
    "\n",
    "        ax[indx].plot(fpr,tpr,label=\"E.S=\"+str(esize)+\", auc=\"+str(roc_auc))\n",
    "        ax[indx].legend(loc=10)\n",
    "        if indx==5:\n",
    "            plt.show()\n",
    "    \n",
    "    prec,recall,_,_ = precision_recall_fscore_support(preds, ytest)\n",
    "    return prec,recall\n",
    "\n",
    "# define ensemble sizes\n",
    "ensemble_size=[10,50,100]\n",
    "avg_prec_rb, avg_rcall_rb=[],[]\n",
    "\n",
    "for esize in ensemble_size:\n",
    "    # lists for storing precisions and recalls\n",
    "    p_rus,r_rus=[],[]\n",
    "    indx=0\n",
    "    fig, ax = plt.subplots(1,5,figsize=(14,3))\n",
    "    \n",
    "    for train_index, test_index in kf.split(features):\n",
    "        X_train, X_test = features[train_index], features[test_index]\n",
    "        y_train, y_test = labels[train_index], labels[test_index]\n",
    "        # RUSBOOST\n",
    "        precision,recall = train_rusboost(X_train, y_train,X_test,y_test, indx, esize)\n",
    "        p_rus.append(precision)\n",
    "        r_rus.append(recall)\n",
    "        indx+=1\n",
    "\n",
    "    print(\"Precision and recall for ensemble size of : \"+str(esize))\n",
    "    avg_prec_rb.append(np.mean(p_rus))\n",
    "    avg_rcall_rb.append(np.mean(r_rus))\n",
    "    print(\"RUSBOOST: Avg Precision= \"+str(round(avg_prec_rb[-1],2))+\" Avg Recall= \"+str(round(avg_rcall_rb[-1],2)))\n",
    "    print(\"=========================================================\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. SMOTE Boost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "class SMOTEBoost():\n",
    "\n",
    "    def __init__(self,instances, labels, n_clf, k_neighbors=3, random_state=76):\n",
    "        self.k = k_neighbors\n",
    "        self.random_state = random_state\n",
    "        self.X = instances\n",
    "        self.y = labels.tolist()\n",
    "        \n",
    "        self.clf = []\n",
    "        self.n_clf = n_clf\n",
    "            \n",
    "        self.weights = []\n",
    "        self.init_weight= 1.0 / len(self.X)\n",
    "        for i in range(len(self.X)):\n",
    "            self.weights.append(self.init_weight)\n",
    "\n",
    "        self.beta = []\n",
    "\n",
    "    def sample(self, n_samples):\n",
    "\n",
    "        S = np.zeros(shape=(n_samples, self.n_features))\n",
    "        # Calculate synthetic samples.\n",
    "        for i in range(n_samples):\n",
    "            # randomly choose a sample from minority class\n",
    "            j = np.random.choice(self.n_minority_samples)\n",
    "            j+= 645 # majority class size\n",
    "\n",
    "            # Find the NN for each sample.\n",
    "            # Exclude the sample itself.\n",
    "            nn = self.neigh.kneighbors(self.X[j].reshape(1, -1),\n",
    "                                       return_distance=False)[:, 1:]\n",
    "            nn_index = np.random.choice(nn[0]) # randomly choose one of its nearest neighbours\n",
    "\n",
    "            dif = self.X[nn_index] - self.X[j]\n",
    "            gap = round(np.random.random(),2)\n",
    "            # create artificial data\n",
    "            S[i, :] = self.X[j, :] + gap * dif[:]\n",
    "\n",
    "        return S\n",
    "\n",
    "    def fit_knn(self,xin):\n",
    "\n",
    "        self.n_minority_samples, self.n_features = xin.shape\n",
    "        # Learn nearest neighbors.\n",
    "        self.neigh = NearestNeighbors(n_neighbors=self.k + 1)\n",
    "        self.neigh.fit(xin)\n",
    "    \n",
    "    def predict(self, instance):\n",
    "        # compute class scores and return the class that has the max score\n",
    "        p_score, n_score = 0 , 0\n",
    "        instance = instance.reshape(1,-1)\n",
    "        for i in range(self.n_clf):\n",
    "            pred = self.clf[i].predict(instance)[0]\n",
    "            probs = self.clf[i].predict_proba(instance)[0]\n",
    "            if pred == 1:\n",
    "                p_score += np.log(1/self.beta[i]) * probs[1]\n",
    "            else:\n",
    "                n_score += np.log(1/self.beta[i]) * probs[0]\n",
    "        if n_score <= p_score:\n",
    "            return 1\n",
    "        else:\n",
    "            return 0\n",
    "        \n",
    "    def fit(self):\n",
    "        \n",
    "        stats_c_ = Counter(self.y)\n",
    "        maj_c_ = max(stats_c_, key=stats_c_.get)\n",
    "        min_c_ = min(stats_c_, key=stats_c_.get)\n",
    "        \n",
    "        self.minority_target = min_c_\n",
    "        self.n_samples=self.y.count(1)-self.y.count(0)\n",
    "        \n",
    "        if self.n_samples<0:\n",
    "            self.n_samples=-self.n_samples\n",
    "        \n",
    "        X_min = self.X[np.where(np.asarray(self.y) == self.minority_target)]\n",
    "        self.fit_knn(X_min) # fitt KNN to minority class data\n",
    "        \n",
    "        for k in range(self.n_clf):\n",
    "            \n",
    "            # SMOTE step.\n",
    "            X_syn = self.sample(self.n_samples)\n",
    "            y_syn = np.full(X_syn.shape[0], fill_value=self.minority_target,\n",
    "                            dtype=np.int64)\n",
    "\n",
    "            # Normalize synthetic sample weights based on current training set.\n",
    "            sample_weight_syn = np.empty(X_syn.shape[0], dtype=np.float64)\n",
    "            sample_weight_syn[:] = 1. / X_syn.shape[0]\n",
    "\n",
    "            # Combine the original and synthetic samples.\n",
    "            X = np.vstack((self.X, X_syn))\n",
    "            y = np.append(self.y, y_syn)\n",
    "\n",
    "            # Combine the weights.\n",
    "            sample_weight = \\\n",
    "                np.append(self.weights, sample_weight_syn).reshape(-1, 1)\n",
    "            sample_weight = \\\n",
    "                np.squeeze(normalize(sample_weight, axis=0, norm='l1'))\n",
    "            # construct base classifier\n",
    "            base_classifier = DT(max_depth=2)\n",
    "            base_classifier.fit(X, y, sample_weight)\n",
    "            self.clf.append(base_classifier)\n",
    "\n",
    "            loss = 0\n",
    "            for x,y,w in zip(self.X, self.y, self.weights):\n",
    "                x = x.reshape(1,-1)\n",
    "                probs = self.clf[k].predict_proba(x)[0]\n",
    "                pred = self.clf[k].predict(x)[0]\n",
    "                if y != pred:\n",
    "                    loss+= w * (1 - probs[y] + probs[pred])\n",
    "            loss *= 0.5\n",
    "            beta = loss/(1-loss)\n",
    "            if math.isnan(beta) or beta == 0:\n",
    "                beta = 1\n",
    "            self.beta.append(beta)\n",
    "            if loss > 0: # update weights if there is a loss\n",
    "                \n",
    "                for i in range(len(self.weights)):\n",
    "                    x = self.X[i]\n",
    "                    x = x.reshape(1,-1)\n",
    "                    probs = self.clf[k].predict_proba(x)[0]\n",
    "                    pred = self.clf[k].predict(x)[0]\n",
    "                    #if self.y[i] != pred:\n",
    "                    self.weights[i] = self.weights[i] * (np.power(beta,(0.5*(1 + probs[y] - probs[pred]))))\n",
    "            \n",
    "            # use total sum to normalize sum of weights to 1.0\n",
    "            sum_weight = sum(self.weights)\n",
    "            self.weights = [w / sum_weight for w in self.weights]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Precision and recall for ensemble size of : 10\n",
      "SMOTEBOOST: Avg Precision= 0.66 Avg Recall= 0.65\n",
      "=========================================================\n",
      "Precision and recall for ensemble size of : 50\n",
      "SMOTEBOOST: Avg Precision= 0.66 Avg Recall= 0.65\n",
      "=========================================================\n",
      "Precision and recall for ensemble size of : 100\n",
      "SMOTEBOOST: Avg Precision= 0.67 Avg Recall= 0.65\n",
      "=========================================================\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def train_smoteboost(xtrain, ytrain, xtest, ytest, indx, ensemble_size=10, plot_roc=False):\n",
    "    \n",
    "    smote = SMOTEBoost(features,labels,ensemble_size)\n",
    "    smote.fit()\n",
    "    \n",
    "    preds=[]\n",
    "    for i in range(xtest.shape[0]):\n",
    "        preds.append(smote.predict(xtest[i]))\n",
    "    if plot_roc:\n",
    "        # plot ROC Curves\n",
    "        fpr, tpr, _ = roc_curve(ytest, preds)\n",
    "        roc_auc = auc(fpr, tpr)\n",
    "\n",
    "        ax[indx].plot(fpr,tpr,label=\"E.S=\"+str(esize)+\", auc=\"+str(roc_auc))\n",
    "        ax[indx].legend(loc=10)\n",
    "        if indx==5:\n",
    "            plt.show()\n",
    "    \n",
    "    prec,recall,_,_ = precision_recall_fscore_support(preds, ytest)\n",
    "    return prec,recall\n",
    "\n",
    "ensemble_size=[10,50,100]\n",
    "avg_prec_sb,avg_rcall_sb=[],[]\n",
    "\n",
    "for esize in ensemble_size:\n",
    "    # lists for storing precisions and recalls\n",
    "    p_smote,r_smote=[],[]\n",
    "    indx=0\n",
    "    fig, ax = plt.subplots(1,5,figsize=(14,3))\n",
    "    \n",
    "    for train_index, test_index in kf.split(features):\n",
    "        X_train, X_test = features[train_index], features[test_index]\n",
    "        y_train, y_test = labels[train_index], labels[test_index]\n",
    "        # SMOTEBOOST\n",
    "        precision,recall = train_smoteboost(X_train, y_train,X_test,y_test,indx, esize)\n",
    "        p_smote.append(precision)\n",
    "        r_smote.append(recall)\n",
    "        indx+=1\n",
    "\n",
    "    print(\"Precision and recall for ensemble size of : \"+str(esize))\n",
    "    avg_prec_sb.append(np.mean(p_smote))\n",
    "    avg_rcall_sb.append(np.mean(r_smote))\n",
    "    print(\"SMOTEBOOST: Avg Precision= \"+str(round(avg_prec_sb[-1],2))+\" Avg Recall= \"+str(round(avg_rcall_sb[-1],2)))\n",
    "    print(\"=========================================================\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. RB-Boost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "class RBboost():\n",
    "    \n",
    "    def __init__(self,instances,labels, n_clf, neighbours=5):\n",
    "        self.X = instances\n",
    "        self.y = labels.tolist()\n",
    "        \n",
    "        self.clf = []\n",
    "        self.n_clf = n_clf\n",
    "            \n",
    "        self.weights = []\n",
    "        self.init_weight= 1.0 / len(self.X)\n",
    "        for i in range(len(self.X)):# assign initial weights\n",
    "            self.weights.append(self.init_weight)\n",
    "\n",
    "        self.beta = []\n",
    "        \n",
    "        self.majority_size = self.y.count(0)\n",
    "        self.minority_size = self.y.count(1)\n",
    "        self.k = neighbours\n",
    "        \n",
    "    def sample(self, n_samples, minority):\n",
    "\n",
    "        S = np.zeros(shape=(n_samples, self.n_features))\n",
    "        # Calculate synthetic samples.\n",
    "        j=0\n",
    "        for i in range(n_samples):\n",
    "            if minority: # choose from minority class data\n",
    "                j = np.random.choice(self.minority_size)\n",
    "                j+= self.majority_size\n",
    "            else: # choose from majority class data\n",
    "                j = np.random.choice(self.majority_size)\n",
    "\n",
    "            # Find the NN for each sample.\n",
    "            # Exclude the sample itself.\n",
    "            nn = self.neigh.kneighbors(self.X[j].reshape(1, -1),\n",
    "                                       return_distance=False)[:, 1:]\n",
    "            \n",
    "            nn_index = np.random.choice(nn[0]) # select one of the nearest neighbours randomly\n",
    "\n",
    "            dif = self.X[nn_index] - self.X[j]\n",
    "            gap = np.random.random()\n",
    "\n",
    "            S[i, :] = self.X[j, :] + gap * dif[:] # create artificial data\n",
    "\n",
    "        return S\n",
    "\n",
    "    def fit_knn(self,xin):\n",
    "\n",
    "        self.n_samples, self.n_features = xin.shape\n",
    "        # Learn nearest neighbors.\n",
    "        self.neigh = NearestNeighbors(n_neighbors=self.k + 1)\n",
    "        self.neigh.fit(xin)\n",
    "\n",
    "    def random_balance(self):\n",
    "        \n",
    "        total_size = len(self.y)\n",
    "        # randomly pick a majority class size > 2\n",
    "        new_majority_size = np.random.randint(2, total_size - 2)\n",
    "        new_minority_size = total_size - new_majority_size\n",
    "        \n",
    "        X_min = self.X[np.where(np.asarray(self.y) == 1)]\n",
    "        X_maj = self.X[np.where(np.asarray(self.y) == 0)]\n",
    "        \n",
    "        \n",
    "        randbal_data=[]\n",
    "        if new_majority_size < self.majority_size:# if new majority class size was lower than original\n",
    "            # randomly select 'new_majority_size' number of instances from original maj class\n",
    "            self.fit_knn(X_min)\n",
    "            indices = np.random.choice(self.majority_size, new_majority_size)\n",
    "            \n",
    "            for index in indices:\n",
    "                randbal_data.append([self.weights[index],self.X[index],0])\n",
    "            # upsample minority class data\n",
    "            upsamples = self.sample(new_minority_size - self.minority_size,minority=True)\n",
    "            \n",
    "            for i in range(upsamples.shape[0]):\n",
    "                randbal_data.append([1./total_size, upsamples[i], 1])\n",
    "        else:# if new majority class size was higher than original\n",
    "            self.fit_knn(X_maj)\n",
    "            # randomly select 'new_minority_size' number of instances from original minority class\n",
    "            indices = np.random.choice(self.minority_size, new_minority_size)\n",
    "            \n",
    "            for index in indices:\n",
    "                randbal_data.append([self.weights[index],self.X[index],1])\n",
    "            # upsample majority class data\n",
    "            upsamples = self.sample(new_majority_size - self.majority_size,minority=False)\n",
    "            for i in range(upsamples.shape[0]):\n",
    "                randbal_data.append([1./total_size, upsamples[i], 1])\n",
    "        return randbal_data\n",
    "    \n",
    "    def fit(self):\n",
    "        for k in range(self.n_clf):\n",
    "\n",
    "            # check tree accuracy , remove if it was below .5\n",
    "            acc = 0.0\n",
    "            acc_sampled = 0.0\n",
    "            base_classifier = None\n",
    "            while acc < 0.5 or acc_sampled < 0.5:\n",
    "                sampled = self.random_balance()\n",
    "\n",
    "                sampled_weight = [s[0] for s in sampled]\n",
    "                sampled_X      = [s[1] for s in sampled]\n",
    "                sampled_Y       = [s[2] for s in sampled]\n",
    "                   \n",
    "                # try again fitting new tree\n",
    "                base_classifier = DT(max_depth=2)\n",
    "                base_classifier.fit(sampled_X, sampled_Y, sampled_weight)\n",
    "            \n",
    "                acc = accuracy_score(self.y, base_classifier.predict(self.X))\n",
    "                acc_sampled = accuracy_score(sampled_Y, base_classifier.predict(sampled_X))\n",
    "            \n",
    "            self.clf.append(base_classifier)\n",
    "            loss = 0\n",
    "            for x,y,w in zip(self.X, self.y, self.weights):\n",
    "                x = x.reshape(1,-1)\n",
    "                probs = self.clf[k].predict_proba(x)[0]\n",
    "                pred = self.clf[k].predict(x)[0]\n",
    "                if y != pred:\n",
    "                    loss+= w * (1 - probs[y] + probs[pred])\n",
    "            loss *= 0.5\n",
    "            beta = loss/(1-loss)\n",
    "            if math.isnan(beta) or beta == 0:\n",
    "                beta = 1\n",
    "            self.beta.append(beta)\n",
    "            if loss > 0: # update weights if there is a loss\n",
    "                \n",
    "                for i in range(len(self.weights)):\n",
    "                    x = self.X[i]\n",
    "                    x = x.reshape(1,-1)\n",
    "                    probs = self.clf[k].predict_proba(x)[0]\n",
    "                    pred = self.clf[k].predict(x)[0]\n",
    "                    #if self.y[i] != pred:\n",
    "                    self.weights[i] = self.weights[i] * (np.power(beta,(0.5*(1 + probs[y] - probs[pred]))))\n",
    "            \n",
    "            # use total sum to normalize sum of weights to 1.0\n",
    "            sum_weight = sum(self.weights)\n",
    "            self.weights = [w / sum_weight for w in self.weights]\n",
    "            \n",
    "    def predict(self, instance):\n",
    "        p_score, n_score = 0 , 0\n",
    "        instance = instance.reshape(1,-1)\n",
    "        for i in range(self.n_clf):\n",
    "            pred = self.clf[i].predict(instance)[0]\n",
    "            probs = self.clf[i].predict_proba(instance)[0]\n",
    "            if pred == 1:\n",
    "                p_score += np.log(1./self.beta[i])* probs[1]\n",
    "            else:\n",
    "                n_score += np.log(1./self.beta[i])* probs[0]\n",
    "        if n_score <= p_score:\n",
    "            return 1\n",
    "        else:\n",
    "            return 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Precision and recall for ensemble size of : 10\n",
      "RBBOOST: Avg Precision= 0.51 Avg Recall= 0.41\n",
      "=========================================================\n",
      "Precision and recall for ensemble size of : 50\n",
      "RBBOOST: Avg Precision= 0.62 Avg Recall= 0.59\n",
      "=========================================================\n",
      "Precision and recall for ensemble size of : 100\n",
      "RBBOOST: Avg Precision= 0.66 Avg Recall= 0.63\n",
      "=========================================================\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def train_rbboost(xtrain, ytrain, xtest, ytest, indx,ensemble_size=10, plot_roc=False):\n",
    "    \n",
    "    rbboost = RBboost(features,labels,ensemble_size)\n",
    "    rbboost.fit()\n",
    "    \n",
    "    preds=[]\n",
    "    for i in range(xtest.shape[0]):\n",
    "        preds.append(rbboost.predict(xtest[i]))\n",
    "    if plot_roc:\n",
    "        # plot ROC curves   \n",
    "        fpr, tpr, _ = roc_curve(ytest, preds)\n",
    "        roc_auc = auc(fpr, tpr)\n",
    "\n",
    "        ax[indx].plot(fpr,tpr,label=\"E.S=\"+str(esize)+\", auc=\"+str(round(roc_auc,2)))\n",
    "        ax[indx].legend(loc=10)\n",
    "        if indx==5:\n",
    "            plt.show()\n",
    "    \n",
    "    prec,recall,_,_ = precision_recall_fscore_support(preds, ytest)\n",
    "    return prec,recall\n",
    "\n",
    "\n",
    "ensemble_size=[10,50,100]\n",
    "avg_prec_rbb,avg_rcall_rbb=[],[]\n",
    "\n",
    "for esize in ensemble_size:\n",
    "    # lists for storing precisions and recalls\n",
    "    p_rbb,r_rbb=[],[]\n",
    "    indx=0\n",
    "    fig, ax = plt.subplots(1,5,figsize=(14,3))\n",
    "    \n",
    "    for train_index, test_index in kf.split(features):\n",
    "        X_train, X_test = features[train_index], features[test_index]\n",
    "        y_train, y_test = labels[train_index], labels[test_index]\n",
    "        # RBBOOST\n",
    "        precision,recall = train_rbboost(X_train, y_train,X_test,y_test, indx, esize)\n",
    "        p_rbb.append(precision)\n",
    "        r_rbb.append(recall)\n",
    "        indx+=1\n",
    "        \n",
    "    print(\"Precision and recall for ensemble size of : \"+str(esize))\n",
    "    avg_prec_rbb.append(np.mean(p_rbb))\n",
    "    avg_rcall_rbb.append(np.mean(r_rbb))\n",
    "    print(\"RBBOOST: Avg Precision= \"+str(round(avg_prec_rbb[-1],2))+\" Avg Recall= \"+str(round(avg_rcall_rbb[-1],2)))\n",
    "    print(\"=========================================================\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 5. SVM and Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Precision and recall for ensemble size of : 10\n",
      "RANDOM Forest: Avg Precision= 0.92 Avg Recall= 0.94\n",
      "=========================================================\n",
      "Precision and recall for ensemble size of : 50\n",
      "RANDOM Forest: Avg Precision= 0.96 Avg Recall= 0.96\n",
      "=========================================================\n",
      "Precision and recall for ensemble size of : 100\n",
      "RANDOM Forest: Avg Precision= 0.96 Avg Recall= 0.96\n",
      "=========================================================\n",
      "SVM: Precision= 1.0 Recall= 0.97\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def train_svm(xtrain, ytrain, xtest, ytest):\n",
    "    clf = LinearSVC(random_state=0)\n",
    "    clf.fit(xtrain,ytrain)\n",
    "    \n",
    "    preds=[]\n",
    "    for i in range(xtest.shape[0]):\n",
    "        preds.append(clf.predict(xtest[i].reshape(1,-1)))\n",
    "    prec,recall,_,_ = precision_recall_fscore_support(preds, ytest)\n",
    "    \n",
    "    return prec,recall\n",
    "\n",
    "def train_randomforest(xtrain, ytrain, xtest, ytest, indx,ensemble_size=10 , plot_roc=False):\n",
    "    clf = RandomForestClassifier(n_estimators=ensemble_size,random_state=0)\n",
    "    clf.fit(xtrain,ytrain)\n",
    "    \n",
    "    preds=[]\n",
    "    for i in range(xtest.shape[0]):\n",
    "        preds.append(clf.predict(xtest[i].reshape(1,-1)))\n",
    "    \n",
    "    if plot_roc:\n",
    "        fpr, tpr, _ = roc_curve(ytest, preds)\n",
    "        roc_auc = auc(fpr, tpr)\n",
    "\n",
    "        ax[indx].plot(fpr,tpr,label=\"Ensemble Size=\"+str(esize)+\", auc=\"+str(roc_auc))\n",
    "        ax[indx].legend(loc=10)\n",
    "        if indx==5:\n",
    "            plt.show()\n",
    "    \n",
    "    prec,recall,_,_ = precision_recall_fscore_support(preds, ytest)\n",
    "    \n",
    "    return prec,recall\n",
    "\n",
    "\n",
    "ensemble_size=[10,50,100]# RandomForest\n",
    "avg_prec_rf, avg_rcall_rf=[], []\n",
    "for esize in ensemble_size:\n",
    "    # lists for storing precisions and recalls\n",
    "    p_rf,r_rf=[],[]\n",
    "    indx=0\n",
    "    fig, ax = plt.subplots(1,5,figsize=(14,3))\n",
    "    for train_index, test_index in kf.split(features):\n",
    "        X_train, X_test = features[train_index], features[test_index]\n",
    "        y_train, y_test = labels[train_index], labels[test_index]\n",
    "        # RANDOM Forest\n",
    "        precision,recall= train_randomforest(X_train, y_train,X_test,y_test, indx, esize, True)\n",
    "        p_rf.append(precision)\n",
    "        r_rf.append(recall)\n",
    "        indx+=1\n",
    "    print(\"Precision and recall for ensemble size of : \"+str(esize))\n",
    "    avg_prec_rf.append(np.mean(p_rf))\n",
    "    avg_rcall_rf.append(np.mean(r_rf))\n",
    "    print(\"RANDOM Forest: Avg Precision= \"+str(round(avg_prec_rf[-1],2))+\" Avg Recall= \"+str(round(avg_rcall_rf[-1],2)))\n",
    "    print(\"=========================================================\")\n",
    "\n",
    "# SVM\n",
    "precision,recall = train_svm(X_train, y_train,X_test,y_test)\n",
    "p_svm = precision\n",
    "r_svm = recall\n",
    "\n",
    "print(\"SVM: Precision= \"+str(round(p_svm[0],2))+\" Recall= \"+str(round(r_svm[0],2)))\n",
    "print()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 6. Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average Classifier Precision for AdaBoost :          0.77\n",
      "Average Classifier Precision for RUSBoost :          0.82\n",
      "Average Classifier Precision for SMOTEBoost :        0.66\n",
      "Average Classifier Precision for RandomBalanceBoost :0.6\n",
      "Average Classifier Precision for RandomForest :      0.95\n",
      "Average Classifier Precision for SVM :               1.0\n",
      "--------------------------------------------------------------\n",
      "*Best performing method based on Average Precision of classifiers: \"SVM\"\n",
      "\n",
      "*Best Performing Ensemble Classifier is \"Random Forset\" Runner up (second best) is RUSBOOST\n",
      "\n",
      "--------------------------------------------------------------------------------\n",
      "\n",
      "Taha Samavati - 98722134 - Analysis of final results\n"
     ]
    }
   ],
   "source": [
    "# Plot and analyse results\n",
    "def autolabel(rects,i):\n",
    "    \"\"\"Attach a text label above each bar \"\"\"\n",
    "    for rect in rects:\n",
    "        height = rect.get_height()\n",
    "        ax[i].annotate('{}'.format(height),\n",
    "                    xy=(rect.get_x() + rect.get_width() / 2, height),\n",
    "                    xytext=(0, 3),  # 3 points vertical offset\n",
    "                    textcoords=\"offset points\",\n",
    "                    ha='center', va='bottom')\n",
    "\n",
    "y_preds = [avg_prec_ada[0], avg_prec_rb[0], avg_prec_sb[0], avg_prec_rbb[0], avg_prec_rf[0],p_svm[0]]\n",
    "y_preds1 = [avg_prec_ada[1], avg_prec_rb[1], avg_prec_sb[1], avg_prec_rbb[1], avg_prec_rf[1],p_svm[0]]\n",
    "y_preds2 = [avg_prec_ada[2], avg_prec_rb[2], avg_prec_sb[2], avg_prec_rbb[2], avg_prec_rf[2],p_svm[0]]\n",
    "\n",
    "y_preds3 = [avg_rcall_ada[0], avg_rcall_rb[0], avg_rcall_sb[0], avg_rcall_rbb[0], avg_rcall_rf[0],r_svm[0]]\n",
    "y_preds4 = [avg_rcall_ada[1], avg_rcall_rb[1], avg_rcall_sb[1], avg_rcall_rbb[1], avg_rcall_rf[1],r_svm[0]]\n",
    "y_preds5 = [avg_rcall_ada[2], avg_rcall_rb[2], avg_rcall_sb[2], avg_rcall_rbb[2], avg_rcall_rf[2],r_svm[0]]\n",
    "\n",
    "method_names=['AdaBoost.M2','RUSBoost','SMOTEBoost','RandomBalanceBoost','RandomForest','SVM']\n",
    "ensemble_size=['Ensemble=10','Ensemble=50','Ensemble=100']\n",
    "fig, ax = plt.subplots(2,figsize=(12,10))\n",
    "labels = method_names\n",
    "\n",
    "data = np.round(y_preds,2)\n",
    "data_1 = np.round(y_preds1,2)\n",
    "data_2 = np.round(y_preds2,2)\n",
    "data_3 = np.round(y_preds3,2)\n",
    "data_4 = np.round(y_preds4,2)\n",
    "data_5 = np.round(y_preds5,2)\n",
    "\n",
    "x = np.arange(len(labels))  # the label locations\n",
    "width = 0.15  # the width of the bars\n",
    "\n",
    "\n",
    "rects1 = ax[0].bar(x - (width+0.3)/2, data, width, label=\"Ensemble_size=10\")\n",
    "rects2 = ax[0].bar(x - (width+0.15)/4, data_1, width, label=\"Ensemble_size=50\")\n",
    "rects3 = ax[0].bar(x + (width+0.15)/4, data_2, width, label=\"Ensemble_size=100\")\n",
    "\n",
    "rects4 = ax[1].bar(x - (width+0.3)/2, data, width, label=\"Ensemble_size=10\")\n",
    "rects5 = ax[1].bar(x - (width+0.15)/4, data_1, width, label=\"Ensemble_size=50\")\n",
    "rects6 = ax[1].bar(x + (width+0.15)/4, data_2, width, label=\"Ensemble_size=100\")\n",
    "\n",
    "# Add some text for labels, title and custom x-axis tick labels, etc.\n",
    "ax[0].set_ylabel('Precision')\n",
    "ax[0].set_title('Precision of each classifier')\n",
    "ax[0].set_xticks(x)\n",
    "ax[0].set_xticklabels(labels)\n",
    "ax[0].legend()\n",
    "\n",
    "ax[1].set_ylabel('Recall')\n",
    "ax[1].set_title('Recall of each classifier')\n",
    "ax[1].set_xticks(x)\n",
    "ax[1].set_xticklabels(labels)\n",
    "ax[1].legend()\n",
    "\n",
    "autolabel(rects1,0)\n",
    "autolabel(rects2,0)\n",
    "autolabel(rects3,0)\n",
    "autolabel(rects4,1)\n",
    "autolabel(rects5,1)\n",
    "autolabel(rects6,1)\n",
    "\n",
    "fig.tight_layout()\n",
    "\n",
    "plt.show()\n",
    "\n",
    "print(\"Average Classifier Precision for AdaBoost :          \" + str(round(np.mean(avg_prec_ada),2)))\n",
    "print(\"Average Classifier Precision for RUSBoost :          \" + str(round(np.mean(avg_prec_rb),2)))\n",
    "print(\"Average Classifier Precision for SMOTEBoost :        \" + str(round(np.mean(avg_prec_sb),2)))\n",
    "print(\"Average Classifier Precision for RandomBalanceBoost :\" + str(round(np.mean(avg_prec_rbb),2)))\n",
    "print(\"Average Classifier Precision for RandomForest :      \" + str(round(np.mean(avg_prec_rf),2)))\n",
    "print(\"Average Classifier Precision for SVM :               \" + str(round(p_svm[0],2)))\n",
    "print(\"--------------------------------------------------------------\")\n",
    "print(\"*Best performing method based on Average Precision of classifiers: \\\"SVM\\\"\" )\n",
    "print()\n",
    "print(\"*Best Performing Ensemble Classifier is \\\"Random Forset\\\" Runner up (second best) is RUSBOOST\" )\n",
    "print()\n",
    "print(\"--------------------------------------------------------------------------------\")\n",
    "print()\n",
    "print(\"Taha Samavati - 98722134 - Analysis of final results\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 7.Bonus === ANOVA Test ==="
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Perform 10 random 30-70 splits and evaluate classifiers on them\n",
    "* store results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration # 1 ...\n",
      "Iteration # 2 ...\n",
      "Iteration # 3 ...\n",
      "Iteration # 4 ...\n",
      "Iteration # 5 ...\n",
      "Iteration # 6 ...\n",
      "Iteration # 7 ...\n",
      "Iteration # 8 ...\n",
      "Iteration # 9 ...\n",
      "Iteration # 10 ...\n"
     ]
    }
   ],
   "source": [
    "esize = 50\n",
    "p_rf,p_svm,p_rbb,p_rb,p_sb,p_ada=[],[],[],[],[],[]\n",
    "for i in range(10):\n",
    "    print(\"Iteration # \"+str(i+1)+\" ...\")\n",
    "    X_train, X_test, y_train, y_test = train_test_split(features, labels, test_size=0.3, random_state=42+i)\n",
    "    prec,_ = train_randomforest(X_train, y_train,X_test,y_test,0 , esize, False)\n",
    "    p_rf.append(prec)\n",
    "    prec,_ = train_rbboost(X_train, y_train,X_test,y_test,0, esize, False)\n",
    "    p_rbb.append(prec)\n",
    "    prec,_ = train_rusboost(X_train, y_train,X_test,y_test,0, esize, False)\n",
    "    p_rb.append(prec)\n",
    "    prec,_ = train_smoteboost(X_train, y_train,X_test,y_test,0, esize, False)\n",
    "    p_sb.append(prec)\n",
    "    prec,_ = train_adaboost(X_train, y_train,X_test,y_test,0, esize, False)\n",
    "    p_ada.append(prec)\n",
    "    prec,_ = train_svm(X_train, y_train,X_test,y_test)\n",
    "    p_svm.append(prec)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## ANOVA Test\n",
    "* Create data frame of classifiers each with their accuracy score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_scores = p_rf + p_rbb + p_rb + p_sb + p_ada + p_svm\n",
    "clf_names = (['RandomForest'] * len(p_rf)) +  (['RandomBalanceBoost'] * len(p_rbb)) +  (['RUSBoost'] * len(p_rb)) +  (['SMOTEBoost'] * len(p_sb)) +  (['AdaBoostM2'] * len(p_ada)) +  (['SVM'] * len(p_svm))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.DataFrame({'classifier': clf_names, 'score': all_scores})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(len(data['score'])):\n",
    "    data['score'].loc[i] = data['score'].loc[i][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "# add new column group mean to data frame\n",
    "ss_residual=0\n",
    "data['group_mean']=0\n",
    "for i in range(data.shape[0]):\n",
    "    if data['classifier'].loc[i] == 'RandomForest':\n",
    "        data['group_mean'].loc[i] = np.mean(p_rf)\n",
    "    if data['classifier'].loc[i] == 'RandomBalanceBoost':\n",
    "        data['group_mean'].loc[i] = np.mean(p_rbb)\n",
    "    if data['classifier'].loc[i] == 'RUSBoost':\n",
    "        data['group_mean'].loc[i] = np.mean(p_rb)\n",
    "    if data['classifier'].loc[i] == 'SMOTEBoost':\n",
    "        data['group_mean'].loc[i] = np.mean(p_sb)\n",
    "    if data['classifier'].loc[i] == 'AdaBoostM2':\n",
    "        data['group_mean'].loc[i] = np.mean(p_ada)\n",
    "    if data['classifier'].loc[i] == 'SVM':\n",
    "        data['group_mean'].loc[i] = np.mean(p_svm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7712471060405698"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# compute overall mean\n",
    "overall_mean = data['score'].mean()\n",
    "overall_mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.7650257087004517"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# compute Total Sum of Squares \n",
    "data['overall_mean'] = overall_mean\n",
    "ss_total = sum((data['score'] - data['overall_mean'])**2)\n",
    "ss_total"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.7540917621642278"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# compute Sum of Squares (Variance) within groups\n",
    "ss_residual = sum((data['score'] - data['group_mean'])**2)\n",
    "ss_residual"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7970251992263431"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# compute Sum of Squares between groups\n",
    "ss_explained = sum((data['group_mean'] - data['overall_mean'])**2)\n",
    "ss_explained"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "df_residual : 54\n",
      "mean square residual (within groups)\n",
      "0.032483180780819035\n"
     ]
    }
   ],
   "source": [
    "# num all observations and number of groups (distinct classifiers)\n",
    "n_groups = len(set(data['classifier']))\n",
    "n_obs = data.shape[0]\n",
    "# degrees of freedom for residual\n",
    "df_residual = n_obs - n_groups\n",
    "print(\"df_residual : \"+str(df_residual))\n",
    "# variance within groups / (num_all_obs - num classifiers)\n",
    "ms_residual = ss_residual / df_residual\n",
    "print(\"mean square residual (within groups)\")\n",
    "print(ms_residual)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "df_explained : 5\n",
      "mean square explained (between groups)\n",
      "0.1594050398452686\n"
     ]
    }
   ],
   "source": [
    "df_explained = n_groups - 1\n",
    "print(\"df_explained : \"+str(df_explained))\n",
    "# compute variance between groups / (n_groups - 1)\n",
    "ms_explained = ss_explained / df_explained\n",
    "print(\"mean square explained (between groups)\")\n",
    "print(ms_explained)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.907310060577428"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# compute F-Value\n",
    "f = ms_explained / ms_residual\n",
    "f"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0008968784502048077"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# compute p-value\n",
    "import scipy.stats\n",
    "p_value = 1 - scipy.stats.f.cdf(f, df_explained, df_residual)\n",
    "p_value"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* p is less than 0.05\n",
    "* Look up the F-distibution table (p<0.05), (df_residual,df_explained)=(54,5)\n",
    "* **F(54,5) = 1.93 < 4.90**\n",
    "* So we can reject the null hypothesis\n",
    "* which means that **there is a meaningfull difference between classifiers performance.**\n",
    "* This was result of 1-way ANOVA Test\n",
    "* according to the result, The best classifier is one with lowes variance within groups and maximum average prec.\n",
    "* the best is SVM - Second best is Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>classifier</th>\n",
       "      <th>score</th>\n",
       "      <th>overall_mean</th>\n",
       "      <th>group_mean</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>1</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.941119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.964824</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.941119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.968586</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.941119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.979167</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.941119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.989899</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.941119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.978142</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.941119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.984536</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.941119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.979487</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.941119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.972678</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.941119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.975248</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.941119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>RandomBalanceBoost</td>\n",
       "      <td>0.486772</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.645148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>RandomBalanceBoost</td>\n",
       "      <td>0.683417</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.645148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>RandomBalanceBoost</td>\n",
       "      <td>0.507853</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.645148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>RandomBalanceBoost</td>\n",
       "      <td>0.416667</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.645148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>RandomBalanceBoost</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.645148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>RandomBalanceBoost</td>\n",
       "      <td>0.530055</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.645148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>RandomBalanceBoost</td>\n",
       "      <td>0.443299</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.645148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>RandomBalanceBoost</td>\n",
       "      <td>0.441026</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.645148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>RandomBalanceBoost</td>\n",
       "      <td>0.672131</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.645148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>RandomBalanceBoost</td>\n",
       "      <td>0.653465</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.645148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>RUSBoost</td>\n",
       "      <td>0.624339</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.807502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>RUSBoost</td>\n",
       "      <td>0.904523</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.807502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>RUSBoost</td>\n",
       "      <td>0.722513</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.807502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>RUSBoost</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.807502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>RUSBoost</td>\n",
       "      <td>0.656566</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.807502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>RUSBoost</td>\n",
       "      <td>0.721311</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.807502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>RUSBoost</td>\n",
       "      <td>0.804124</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.807502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>RUSBoost</td>\n",
       "      <td>0.892308</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.807502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>RUSBoost</td>\n",
       "      <td>0.896175</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.807502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>RUSBoost</td>\n",
       "      <td>0.841584</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.807502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>SMOTEBoost</td>\n",
       "      <td>0.275132</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.664540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>SMOTEBoost</td>\n",
       "      <td>0.341709</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.664540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>SMOTEBoost</td>\n",
       "      <td>0.340314</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.664540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>SMOTEBoost</td>\n",
       "      <td>0.364583</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.664540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>SMOTEBoost</td>\n",
       "      <td>0.388889</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.664540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>SMOTEBoost</td>\n",
       "      <td>0.36612</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.664540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>SMOTEBoost</td>\n",
       "      <td>0.365979</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.664540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>SMOTEBoost</td>\n",
       "      <td>0.374359</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.664540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>SMOTEBoost</td>\n",
       "      <td>0.300546</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.664540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>SMOTEBoost</td>\n",
       "      <td>0.39604</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.664540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>AdaBoostM2</td>\n",
       "      <td>0.984127</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.783312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>AdaBoostM2</td>\n",
       "      <td>0.984925</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.783312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>AdaBoostM2</td>\n",
       "      <td>0.994764</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.783312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>AdaBoostM2</td>\n",
       "      <td>0.994792</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.783312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>AdaBoostM2</td>\n",
       "      <td>0.979798</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.783312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>AdaBoostM2</td>\n",
       "      <td>0.989071</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.783312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>AdaBoostM2</td>\n",
       "      <td>0.979381</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.783312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>AdaBoostM2</td>\n",
       "      <td>0.994872</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.783312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>AdaBoostM2</td>\n",
       "      <td>0.972678</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.783312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>AdaBoostM2</td>\n",
       "      <td>0.985149</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.783312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>SVM</td>\n",
       "      <td>0.994709</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.919904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>SVM</td>\n",
       "      <td>0.939698</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.919904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>SVM</td>\n",
       "      <td>1</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.919904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>SVM</td>\n",
       "      <td>0.96875</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.919904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>SVM</td>\n",
       "      <td>1</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.919904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>SVM</td>\n",
       "      <td>0.961749</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.919904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>SVM</td>\n",
       "      <td>0.969072</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.919904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>SVM</td>\n",
       "      <td>0.984615</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.919904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>SVM</td>\n",
       "      <td>0.939891</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.919904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>SVM</td>\n",
       "      <td>0.980198</td>\n",
       "      <td>0.771247</td>\n",
       "      <td>0.919904</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            classifier     score  overall_mean  group_mean\n",
       "0         RandomForest         1      0.771247    0.941119\n",
       "1         RandomForest  0.964824      0.771247    0.941119\n",
       "2         RandomForest  0.968586      0.771247    0.941119\n",
       "3         RandomForest  0.979167      0.771247    0.941119\n",
       "4         RandomForest  0.989899      0.771247    0.941119\n",
       "5         RandomForest  0.978142      0.771247    0.941119\n",
       "6         RandomForest  0.984536      0.771247    0.941119\n",
       "7         RandomForest  0.979487      0.771247    0.941119\n",
       "8         RandomForest  0.972678      0.771247    0.941119\n",
       "9         RandomForest  0.975248      0.771247    0.941119\n",
       "10  RandomBalanceBoost  0.486772      0.771247    0.645148\n",
       "11  RandomBalanceBoost  0.683417      0.771247    0.645148\n",
       "12  RandomBalanceBoost  0.507853      0.771247    0.645148\n",
       "13  RandomBalanceBoost  0.416667      0.771247    0.645148\n",
       "14  RandomBalanceBoost  0.722222      0.771247    0.645148\n",
       "15  RandomBalanceBoost  0.530055      0.771247    0.645148\n",
       "16  RandomBalanceBoost  0.443299      0.771247    0.645148\n",
       "17  RandomBalanceBoost  0.441026      0.771247    0.645148\n",
       "18  RandomBalanceBoost  0.672131      0.771247    0.645148\n",
       "19  RandomBalanceBoost  0.653465      0.771247    0.645148\n",
       "20            RUSBoost  0.624339      0.771247    0.807502\n",
       "21            RUSBoost  0.904523      0.771247    0.807502\n",
       "22            RUSBoost  0.722513      0.771247    0.807502\n",
       "23            RUSBoost      0.75      0.771247    0.807502\n",
       "24            RUSBoost  0.656566      0.771247    0.807502\n",
       "25            RUSBoost  0.721311      0.771247    0.807502\n",
       "26            RUSBoost  0.804124      0.771247    0.807502\n",
       "27            RUSBoost  0.892308      0.771247    0.807502\n",
       "28            RUSBoost  0.896175      0.771247    0.807502\n",
       "29            RUSBoost  0.841584      0.771247    0.807502\n",
       "30          SMOTEBoost  0.275132      0.771247    0.664540\n",
       "31          SMOTEBoost  0.341709      0.771247    0.664540\n",
       "32          SMOTEBoost  0.340314      0.771247    0.664540\n",
       "33          SMOTEBoost  0.364583      0.771247    0.664540\n",
       "34          SMOTEBoost  0.388889      0.771247    0.664540\n",
       "35          SMOTEBoost   0.36612      0.771247    0.664540\n",
       "36          SMOTEBoost  0.365979      0.771247    0.664540\n",
       "37          SMOTEBoost  0.374359      0.771247    0.664540\n",
       "38          SMOTEBoost  0.300546      0.771247    0.664540\n",
       "39          SMOTEBoost   0.39604      0.771247    0.664540\n",
       "40          AdaBoostM2  0.984127      0.771247    0.783312\n",
       "41          AdaBoostM2  0.984925      0.771247    0.783312\n",
       "42          AdaBoostM2  0.994764      0.771247    0.783312\n",
       "43          AdaBoostM2  0.994792      0.771247    0.783312\n",
       "44          AdaBoostM2  0.979798      0.771247    0.783312\n",
       "45          AdaBoostM2  0.989071      0.771247    0.783312\n",
       "46          AdaBoostM2  0.979381      0.771247    0.783312\n",
       "47          AdaBoostM2  0.994872      0.771247    0.783312\n",
       "48          AdaBoostM2  0.972678      0.771247    0.783312\n",
       "49          AdaBoostM2  0.985149      0.771247    0.783312\n",
       "50                 SVM  0.994709      0.771247    0.919904\n",
       "51                 SVM  0.939698      0.771247    0.919904\n",
       "52                 SVM         1      0.771247    0.919904\n",
       "53                 SVM   0.96875      0.771247    0.919904\n",
       "54                 SVM         1      0.771247    0.919904\n",
       "55                 SVM  0.961749      0.771247    0.919904\n",
       "56                 SVM  0.969072      0.771247    0.919904\n",
       "57                 SVM  0.984615      0.771247    0.919904\n",
       "58                 SVM  0.939891      0.771247    0.919904\n",
       "59                 SVM  0.980198      0.771247    0.919904"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Taha Samavati - 98722134\n",
    "### Pattern Recognition - Second Project"
   ]
  }
 ],
 "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}