逻辑回归

代码实现如下：

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
def loadDataSet():          #数据的读取，从txt文档中读入          
    dataMat = []   
    labelMat = []
    fr = open('testSet.txt')
    for line in fr.readlines():
        lineArr = line.strip().split() #split将一个字符串进行切片，默认分隔符为空字符
        if(len(lineArr)==0):  #line从文档头开始一行行读取，必须加上判断条件在读到末尾的下一行时跳出，否则会出现列表越界！
            break
        dataMat.append([1.0, float(lineArr[0]), float(lineArr[1])]) #特征值x0,x1,x2，其中x0为偏移量设为1
        labelMat.append(int(lineArr[2]))                            #真实值（标记，分为0和1）
    return dataMat, labelMat

def sigmoid(z):                         
    return 1.0/(1+np.exp(-z))                                           
#返回最佳回归系数theta
def gradAscent(dataMatIn, classLabels):
    #将得到的列表转换为numpy矩阵类型
    dataMatrix = np.mat(dataMatIn) 
    labelMat = np.mat(classLabels).transpose()
    # m->样本数 n->特征数
    m= dataMatrix.shape[0]
    n=dataMatrix.shape[1]
    alpha = 0.001   #步长，学习率
    maxCycles = 500 #限制迭代次数
    #初始化theta参数，每个维度均为1.0
    theta = np.ones((n,1))
    for i in range(maxCycles):
        h = sigmoid(dataMatrix * theta)  #得到h(x)
        error = (h-labelMat)             #得到h(x)-y
        theta = theta - alpha * dataMatrix.transpose() *error  
        #dataMatrix.transpose()*error是代价函数的偏导,返回得到的特征参数θ0，θ1，θ2
    return np.array(theta)


def plotBestFit(theta):              
    dataMat, labelMat = loadDataSet()
    dataArr = np.array(dataMat)
    # n->样本数
    n = dataArr.shape[0]
    #xcord1,ycord1代表正例特征
    #xcord2,ycord2代表负例特征
    xcord1 = []; ycord1 = []
    xcord2 = []; ycord2 = []
    for i in range(n):             #将数据分为正负集两类
        if int(labelMat[i]) == 1:
            xcord1.append(dataArr[i,1]); ycord1.append(dataArr[i, 2])
        else:
            xcord2.append(dataArr[i,1]); ycord2.append(dataArr[i, 2])
    fig = plt.figure()
    ax = fig.add_subplot(111)
    ax.scatter(xcord1, ycord1, s = 30, c = 'r', marker = 'o')  #画出各自的散点图
    ax.scatter(xcord2, ycord2, s = 30, c = 'b',marker='x')
    #设定边界直线x和y的值
    x = range(-3,3)
    """
    决策边界直线方程式
    w0*x0+w1*x1+w2*x2=f(x)
    f(x)=0是两个类别的分界处
    所以： w0+w1*x+w2*y=0 => y = (-w0-w1*x)/w2    
    """
    y = (-theta[0] - theta[1] * x) / theta[2]
    ax.plot(x, y)
    plt.xlabel('X1')
    plt.ylabel('X2')
    plt.show()
    
dataMat,labelMat=loadDataSet()
theta=gradAscent(dataMat,labelMat)
print(theta)
theta=theta.tolist()
mytheta=np.array([[theta[0][0]],[theta[1][0]],[theta[2][0]]])
plotBestFit(mytheta)

输出结果：