SVM/svm.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Support Vector Machines\n",
    "\n",
    "support vector machines (SVMs) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "from matplotlib.colors import ListedColormap\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn import datasets, svm\n",
    "from sklearn.metrics import accuracy_score, classification_report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "np.random.seed(0)\n",
    "\n",
    "iris = datasets.load_iris()\n",
    "X, y = iris.data, iris.target\n",
    "\n",
    "indices = np.random.permutation(len(X))\n",
    "\n",
    "X_train = X[indices[:-30]]\n",
    "y_train = y[indices[:-30]]\n",
    "\n",
    "X_test = X[indices[-30:]]\n",
    "y_test = y[indices[-30:]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf = svm.SVC()\n",
    "clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['virginica']\n"
     ]
    }
   ],
   "source": [
    "result = clf.predict([[3, 5, 4, 2],])\n",
    "\n",
    "print(iris.target_names[result])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['setosa']\n"
     ]
    }
   ],
   "source": [
    "result = clf.predict([[5.1, 3.4, 1.2, 0.1],])\n",
    "\n",
    "print(iris.target_names[result])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.933333333333\n"
     ]
    }
   ],
   "source": [
    "y_pred = clf.predict(X_test)\n",
    "\n",
    "print(accuracy_score(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "     setosa       1.00      1.00      1.00        10\n",
      " versicolor       1.00      0.82      0.90        11\n",
      "  virginica       0.82      1.00      0.90         9\n",
      "\n",
      "avg / total       0.95      0.93      0.93        30\n",
      "\n"
     ]
    }
   ],
   "source": [
    "target_names = iris.target_names\n",
    "print(classification_report(y_test, y_pred, target_names=target_names))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X = iris.data[:, :2] # take only two features\n",
    "y = iris.target\n",
    "h = .02 # step size in the mesh"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "cmap_light = ListedColormap(['#FFAAAA', '#AAFFAA', '#AAAAFF'])\n",
    "cmap_bold = ListedColormap(['#FF0000', '#00FF00', '#0000FF'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape=None, degree=3, gamma='auto', kernel='rbf',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf = svm.SVC()\n",
    "clf.fit(X, y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "decision boundary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Plot the decision boundary. For that, we will assign a color to each\n",
    "# point in the mesh [x_min, m_max]x[y_min, y_max].\n",
    "x_min, x_max = X[:, 0].min() - 1, X[:, 0].max() + 1\n",
    "y_min, y_max = X[:, 1].min() - 1, X[:, 1].max() + 1\n",
    "xx, yy = np.meshgrid(\n",
    "    np.arange(x_min, x_max, h),\n",
    "    np.arange(y_min, y_max, h)\n",
    ")\n",
    "Z = clf.predict(np.c_[xx.ravel(), yy.ravel()])\n",
    "\n",
    "Z = Z.reshape(xx.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXEAAAEKCAYAAADkYmWmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n0yKulFIxTIu4UkrFsP8HyNExZYp277cAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1182673c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "plt.pcolormesh(xx, yy, Z, cmap=cmap_light)\n",
    "plt.scatter(X[:, 0], X[:, 1], c=y, cmap=cmap_bold)\n",
    "plt.xlim(xx.min(), xx.max())\n",
    "plt.ylim(yy.min(), yy.max())\n",
    "plt.title('SVM Classification Results')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [Root]",
   "language": "python",
   "name": "Python [Root]"
  },
  "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
}