机器学习——感知机学习

本篇博客是实现《统计学习方法》中的第二章所讲述的感知机学习 ，这是一个很简单的模型，下面给出算法，

下面给出python代码，包括画出图像

#!/usr/bin/env python
# encoding: utf-8

import matplotlib.pyplot as plt
import numpy as np

x_list = [[3,3],[4,3],[1,2],[1,1],[2,1]]#数据
y_label = [1,1,1,-1,-1]#标签，1为正，-1为负

def perceptron(x_list,y_label,w,b,step):#感知机算法，给出w和b初始值，step为步长
    i = 1
    temp = 1
    length = len(y_label)
    while(bool(i)):#知道没有误分点为止
        index = i%(length+1) - 1
        x = np.array(x_list[index]).T#步骤1，选取数据，并矩阵化
        w = np.array(w).T
        flag = y_label[index]*(np.inner(x,w)+b)#判断是否为误分点
        if flag<= 0 :#是误分点，更新w 和 b
            w = w + step* y_label[index]*x
            b = b + step* y_label[index]
            temp = 0
        i += 1
        if index == length-1 and temp == 1:#如果没有误分点，跳出循环
            return w,b
            break
        temp = 1

def drawline(plt,w,b):#画直线
   x = np.arange(0.0,5.0,0.01)
   y = (-b-w[0]*x)/w[1]
   plt.plot(x,y)

plt.axis([0,5,0,5])#设置x和y轴区间
for x in x_list:#画数据点
    plt.plot(x[0],x[1],'yo-')
plt.title('perceptron')
plt.xlabel('x axis')
plt.ylabel('y axis')
w,b = perceptron(x_list,y_label,[0,0],0,1)
drawline(plt,w,b)
plt.show()

算法很简单，同时算法中给出了相关的注释，应该不难理解，下面给出结果图

下面给出3D的版本

<span style="font-size:12px;">#!/usr/bin/env python
# encoding: utf-8

import matplotlib.pyplot as plt
import numpy as np
from matplotlib import cm
from mpl_toolkits.mplot3d import Axes3D

x_list = [[3,3,2],[4,3,1],[1,2,3],[1,1,2],[2,1,2]]
y_label = [1,1,1,-1,-1]

def perceptron(x_list,y_label,w,b,step):
    i = 1
    temp = 1
    length = len(y_label)
    while(bool(i)):
        index = i%(length+1) - 1
        x = np.array(x_list[index]).T
        w = np.array(w).T
        flag = y_label[index]*(np.inner(x,w)+b)
        if flag<= 0 :
            w = w + step* y_label[index]*x
            b = b + step* y_label[index]
            temp = 0
            print w,b
        i += 1
        if index == length-1 and temp == 1:
            print w,b
            return w,b
            break
        temp = 1

def drawline(plt,w,b):
   x = np.arange(0.0,3.0,0.1)
   y = np.arange(0.0,3.0,0.1)
   x,y = np.meshgrid(x,y)
   z = (-b-w[0]*x-w[1]*x)/w[2]
   ax.plot_surface(x,y,z, rstride=1, cstride=1, cmap=cm.coolwarm, linewidth=0, antialiased=False)

fig = plt.figure()
ax = fig.add_subplot(1,1,1,projection='3d')
for x in x_list:
    ax.scatter(x[0],x[1],x[2],c='b')
w,b = perceptron(x_list,y_label,[0,0,0],0,1)
drawline(plt,w,b)
plt.show()</span>

结果如下