SupportVectorMachine.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "% matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# irisデータセットの読み込み\n",
    "# 品種: SetosaとVersicolour（線形分離可能）\n",
    "# 変数: petal lengthとpetal width\n",
    "from sklearn.datasets import load_iris\n",
    "iris = load_iris()\n",
    "X_1 = iris.data[:100,[2,3]]\n",
    "y_1 = iris.target[:100]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# データの標準化\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "scaler = StandardScaler()\n",
    "scaler.fit(X_1)\n",
    "X_1_std = scaler.transform(X_1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 散布図を出力\n",
    "plt.scatter(X_1_std[:50, [0]], X_1_std[:50, [1]], color='red', marker='s')\n",
    "plt.scatter(X_1_std[50:100, [0]], X_1_std[50:100, [1]], color='blue', marker='x')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ホールド・アウト法による分割\n",
    "from sklearn.model_selection import train_test_split\n",
    "X_1_train, X_1_test, y_1_train, y_1_test = train_test_split(X_1_std, y_1, test_size=0.3, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma='auto', kernel='linear',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 線形SVCの学習\n",
    "from sklearn.svm import SVC\n",
    "svc = SVC(kernel='linear')\n",
    "svc.fit(X_1_train, y_1_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ２次元データの決定曲線をプロットする関数\n",
    "def plot_decision_regions(X, y, classifier, resolution=0.02):\n",
    "    from matplotlib.colors import ListedColormap\n",
    "    markers = (\"s\", \"x\", \"o\", \"^\", \"v\")\n",
    "    colors = (\"red\", \"blue\", \"lightgreen\", \"gray\", \"cyan\")\n",
    "    cmap = ListedColormap(colors[:len(np.unique(y))])\n",
    "    \n",
    "    x1_min, x1_max = X[:, 0].min()-1 , X[:, 0].max() +1\n",
    "    x2_min, x2_max = X[:, 1].min()-1 , X[:, 1].max() +1\n",
    "    \n",
    "    xx1, xx2 = np.meshgrid(np.arange(x1_min, x1_max, resolution), np.arange(x2_min, x2_max, resolution))\n",
    "    Z = classifier.predict(np.array([xx1.ravel(), xx2.ravel()]).T)\n",
    "    Z = Z.reshape(xx1.shape)\n",
    "    plt.contourf(xx1, xx2, Z, alpha=0.4, cmap=cmap)\n",
    "    plt.xlim(xx1.min(), xx1.max())\n",
    "    plt.ylim(xx2.min(), xx2.max())\n",
    "    for idx, cl in enumerate(np.unique(y)):\n",
    "        plt.scatter(x=X[y == cl, 0], y=X[y == cl, 1], alpha=0.8, c=cmap(idx),marker=markers[idx], label=cl)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 決定曲線をプロットする関数\n",
    "plot_decision_regions(X_1_std, y_1, svc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 正答率を出力\n",
    "svc.score(X_1_test, y_1_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# irisデータセットの読み込み\n",
    "# 品種: VersicolourとVirsiniga（線形分離不可能①）\n",
    "# 変数: petal lengthとpetal width\n",
    "from sklearn.datasets import load_iris\n",
    "iris = load_iris()\n",
    "X_2 = iris.data[50:150,[2,3]]\n",
    "y_2 = iris.target[50:150]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# データの標準化\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "scaler_2 = StandardScaler()\n",
    "scaler_2.fit(X_2)\n",
    "X_2_std = scaler.transform(X_2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 散布図を出力\n",
    "plt.scatter(X_2_std[:50, [0]], X_2_std[:50, [1]], color='red', marker='s')\n",
    "plt.scatter(X_2_std[50:100, [0]], X_2_std[50:100, [1]], color='blue', marker='x')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ホールド・アウト法による分割\n",
    "from sklearn.model_selection import train_test_split\n",
    "X_2_train, X_2_test, y_2_train, y_2_test = train_test_split(X_2_std, y_2, test_size=0.3, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma='auto', kernel='linear',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 線形SVCの学習\n",
    "from sklearn.svm import SVC\n",
    "svc_slack = SVC(kernel='linear', C=1.0)\n",
    "svc_slack.fit(X_2_train, y_2_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 決定境界の出力\n",
    "plot_decision_regions(X_2_std, y_2, svc_slack)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9666666666666667"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 正答率を出力\n",
    "svc_slack.score(X_2_test, y_2_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# XORのデータの作成\n",
    "np.random.seed(0)\n",
    "X_xor = np.random.randn(200, 2)\n",
    "y_xor = np.logical_xor(X_xor[:, 0] > 0, X_xor[:, 1] > 0)\n",
    "y_xor = np.where(y_xor, 1, -1)\n",
    "\n",
    "# データのプロット\n",
    "plt.scatter(X_xor[y_xor == 1, 0], X_xor[y_xor == 1, 1], c='b', marker='x', label='1')\n",
    "plt.scatter(X_xor[y_xor == -1, 0], X_xor[y_xor == -1, 1], c='r', marker='s', label='-1')\n",
    "plt.xlim([-3, 3])\n",
    "plt.ylim([-3, 3])\n",
    "plt.legend(loc='best') # 右上に凡例を出力\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ホールド・アウト法による分割\n",
    "from sklearn.model_selection import train_test_split\n",
    "X_xor_train, X_xor_test, y_xor_train, y_xor_test = train_test_split(X_xor, y_xor, test_size=0.3, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=0.1, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma='auto', kernel='linear',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 線形SVMによる学習\n",
    "linear_svm = SVC(kernel='linear', C=0.1)\n",
    "linear_svm.fit(X_xor_train, y_xor_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 決定曲線のプロット\n",
    "plot_decision_regions(X_xor, y_xor, classifier=linear_svm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6166666666666667"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 正答率を出力\n",
    "linear_svm.score(X_xor_test, y_xor_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=10, cache_size=200, class_weight=None, coef0=0.0,\n",
       "  decision_function_shape='ovr', degree=3, gamma=0.1, kernel='rbf',\n",
       "  max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "  tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# カーネルSVMによる学習\n",
    "# C: 誤分類への厳しさ\n",
    "# gamma: 決定曲線の複雑さ\n",
    "rbf_svm = SVC(kernel='rbf', gamma=0.1, C=10)\n",
    "rbf_svm.fit(X_xor, y_xor)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 決定曲線のプロット\n",
    "plot_decision_regions(X_xor, y_xor, classifier=rbf_svm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9666666666666667"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 正答率を出力\n",
    "rbf_svm.score(X_xor_test, y_xor_test)"
   ]
  }
 ],
 "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}