4.1. K-Nearest Neighbor Classifier .ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " # KNN Classifier\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from src.data import Data\n",
    "from src.plothelp import *\n",
    "import time\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn import model_selection\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import classification_report\n",
    "import numpy as np\n",
    "%matplotlib inline  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data = Data(\"img\")\n",
    "data.load_pickle(\"100x100.pickle \")\n",
    "X_train, X_test, y_train, y_test = data.train_test_split()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(4698, 30000) (522, 30000) (4698,) (522,)\n"
     ]
    }
   ],
   "source": [
    "X_trains = np.asarray(X_train).reshape((len(X_train), -1))\n",
    "X_tests = np.asarray(X_test).reshape((len(X_test), -1))\n",
    "y_trains = np.asarray(y_train)\n",
    "y_tests = np.asarray(y_test)\n",
    "print(X_trains.shape, X_tests.shape, y_trains.shape, y_tests.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classifier train in  49.92044687271118 seconds\n"
     ]
    }
   ],
   "source": [
    "start = time.time()\n",
    "neigh = KNeighborsClassifier(n_neighbors=18)\n",
    "neigh.fit(X_trains,  y_trains)\n",
    "end = time.time()\n",
    "elapsed = end - start\n",
    "print(\"Classifier train in \", elapsed, \"seconds\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classifier tested in  158.88989186286926 seconds\n"
     ]
    }
   ],
   "source": [
    "start = time.time()\n",
    "y =neigh.predict(X_tests)\n",
    "end = time.time()\n",
    "elapsed = end - start\n",
    "print(\"Classifier tested in \", elapsed, \"seconds\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "K-NN 45.5938697318 %\n",
      "Confusion matrix, without normalization\n",
      "[[60 25 48  6]\n",
      " [46 77 47 20]\n",
      " [ 0  0  3  1]\n",
      " [33 25 33 98]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVgAAAEmCAYAAAAnRIjxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1b8a48a6048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "accuracy = accuracy_score(y, y_tests)\n",
    "print(\"K-NN\", accuracy * 100, \"%\")\n",
    "matrix = confusion_matrix(y,y_tests)\n",
    "plot_confusion_matrix(matrix,neigh.classes_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Search for the best Hyper paremeters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classifier tested in  1513.4732248783112 seconds\n"
     ]
    }
   ],
   "source": [
    "parameter_grid = {\"n_neighbors\": np.arange(1,31,2), \"weights\": [\"uniform\", \"distance\"],\n",
    "              \"metric\": [\"euclidean\", \"cityblock\"]}\n",
    "\n",
    "start = time.time()\n",
    "\n",
    "search = model_selection.GridSearchCV(neigh, parameter_grid, n_jobs=4) #run 4 jobs in paralel\n",
    "search.fit(X_trains,y_trains)\n",
    "\n",
    "end = time.time()\n",
    "elapsed = end - start\n",
    "print(\"Classifier tested in \", elapsed, \"seconds\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='euclidean',\n",
      "           metric_params=None, n_jobs=1, n_neighbors=8, p=2,\n",
      "           weights='uniform')\n",
      "0.45785440613\n",
      "{'n_neighbors': 8, 'metric': 'euclidean', 'weights': 'uniform'}\n"
     ]
    }
   ],
   "source": [
    "print(search.best_estimator_)\n",
    "print(search.best_score_)\n",
    "print(search.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y_pred = search.best_estimator_.predict(X_tests)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[ 53   7  19   5]\n",
      " [ 47 101  45  12]\n",
      " [  8   1  37   5]\n",
      " [ 31  18  30 103]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV4AAAEmCAYAAAAqWvi2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nafb/pgzO5SJsRYwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1b8ae1275c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "matrix = confusion_matrix(y_pred,y_tests)\n",
    "plot_confusion_matrix(matrix,search.best_estimator_.classes_)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                   precision    recall  f1-score   support\n",
      "\n",
      "   Fish and Chips       0.38      0.63      0.48        84\n",
      "            Pizza       0.80      0.49      0.61       205\n",
      "            Salad       0.28      0.73      0.41        51\n",
      "Vanilla Ice Cream       0.82      0.57      0.67       182\n",
      "\n",
      "      avg / total       0.69      0.56      0.59       522\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_pred, y_tests))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Data accuracy  score 56.3218390805 %\n"
     ]
    }
   ],
   "source": [
    "print(\"Test Data accuracy  score\", accuracy_score(y_pred, y_tests)*100,\"%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%aimport src.plothelp\n",
    "src.plothelp.plot_misclassified"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [default]",
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}