线性判别分析LDA.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 线性判别分析"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 线性判别分析就是将样例投影到一条直线上，使得同类的样例投影点尽可能小，不同类投影点尽可能远离"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import os  \n",
    "import sys  \n",
    "import numpy as np  \n",
    "import operator  \n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "group1 = np.random.random((8,2))*5+20\n",
    "group2 = np.random.random((8,2))*5+2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "X1 = group1\n",
    "y1 = np.ones((8,1))\n",
    "X0 = group2\n",
    "y0 = np.zeros((8,1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x114137f28>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD8CAYAAABn919SAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\njmIAAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x114060160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X1[:,0],X1[:,1],c='r')\n",
    "plt.scatter(X0[:,0],X0[:,1],c='g')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "mean1=np.array([np.mean(X1[:,0]),np.mean(X1[:,1])])\n",
    "mean0=np.array([np.mean(X0[:,0]),np.mean(X0[:,1])])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 根据周志华西瓜书的61页进行求解"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "from numpy import mat\n",
    "m1=np.shape(X1)[0]\n",
    "sw=np.zeros(shape=(2,2))\n",
    "for i in range(m1):  \n",
    "    xsmean=mat(X1[i,:]-mean1)  \n",
    "    sw+=xsmean.transpose()*xsmean\n",
    "m0=np.shape(X0)[0]  \n",
    "for i in range(m0):  \n",
    "    xsmean=mat(X0[i,:]-mean0)  \n",
    "    sw+=xsmean.transpose()*xsmean\n",
    "w=(mean0-mean1)*(mat(sw).I)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "matrix([[-0.5934389 , -0.57983378]])"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1148fac50>]"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD8CAYAAAB0IB+mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nBILgA0AgCD4ABILgA0AgCD4ABOL/AUbbqB4NN3cqAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113bda198>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X1[:,0],X1[:,1],c='r')\n",
    "plt.scatter(X0[:,0],X0[:,1],c='g')\n",
    "x=np.arange(0,25,0.1)\n",
    "y=np.array((-w[0,0]*x)/w[0,1])\n",
    "plt.plot(x,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}