logistic回归详解（四）：梯度下降训练逻辑回归python实现

项目github地址：bitcarmanlee easy-algorithm-interview-and-practice
欢迎大家star，留言，一起学习进步

1.逻辑回归梯度下降的迭代公式

在参考文献1中，我们推导出了逻辑回归的参数迭代公式为:
$$\theta =\theta -\alpha \ast X^T\ast E$$
接下来我们按照这个思路，用python来手动实现以下逻辑回归算法。

2.准备数据

-0.017612   14.053064   0
-1.395634   4.662541    1
-0.752157   6.538620 0
-1.322371   7.152853    0
0.423363 11.054677   0
0.406704    7.067335    1
0.667394    12.741452   0
-2.460150   6.866805    1
0.569411    9.548755    0
-0.026632   10.427743   0
0.850433    6.920334    1
1.347183    13.175500   0
1.176813    3.167020    1
-1.781871   9.097953    0
-0.566606   5.749003    1
0.931635    1.589505    1
-0.024205   6.151823    1
-0.036453   2.690988    1
-0.196949   0.444165    1
1.014459    5.754399    1
1.985298    3.230619    1
-1.693453   -0.557540   1
-0.576525   11.778922   0
-0.346811   -1.678730   1
-2.124484   2.672471    1
1.217916    9.597015    0
-0.733928   9.098687    0
-3.642001   -1.618087   1
0.315985    3.523953    1
1.416614    9.619232    0
-0.386323   3.989286    1
0.556921    8.294984    1
1.224863    11.587360   0
-1.347803   -2.406051   1
1.196604    4.951851    1
0.275221    9.543647    0
0.470575    9.332488    0
-1.889567   9.542662    0
-1.527893   12.150579   0
-1.185247   11.309318   0
-0.445678   3.297303    1
1.042222    6.105155    1
-0.618787   10.320986   0
1.152083    0.548467    1
0.828534    2.676045    1
-1.237728   10.549033   0
-0.683565   -2.166125   1
0.229456    5.921938    1
-0.959885   11.555336   0
0.492911    10.993324   0
0.184992    8.721488    0
-0.355715   10.325976   0
-0.397822   8.058397    0
0.824839    13.730343   0
1.507278    5.027866    1
0.099671    6.835839    1
-0.344008   10.717485   0
1.785928    7.718645    1
-0.918801   11.560217   0
-0.364009   4.747300    1
-0.841722   4.119083    1
0.490426    1.960539    1
-0.007194   9.075792    0
0.356107    12.447863   0
0.342578    12.281162   0
-0.810823   -1.466018   1
2.530777    6.476801    1
1.296683    11.607559   0
0.475487    12.040035   0
-0.783277   11.009725   0
0.074798    11.023650   0
-1.337472   0.468339    1
-0.102781   13.763651   0
-0.147324   2.874846    1
0.518389    9.887035    0
1.015399    7.571882    0
-1.658086   -0.027255   1
1.319944    2.171228    1
2.056216    5.019981    1
-0.851633   4.375691    1
-1.510047   6.061992    0
-1.076637   -3.181888   1
1.821096    10.283990   0
3.010150    8.401766    1
-1.099458   1.688274    1
-0.834872   -1.733869   1
-0.846637   3.849075    1
1.400102    12.628781   0
1.752842    5.468166    1
0.078557    0.059736    1
0.089392    -0.715300   1
1.825662    12.693808   0
0.197445    9.744638    0
0.126117    0.922311    1
-0.679797   1.220530    1
0.677983    2.556666    1
0.761349    10.693862   0
-2.168791   0.143632    1
1.388610    9.341997    0
0.317029    14.739025   0

原始数据为文本，共三列，其中第一列第二列分别为两个特征，第三列为类别。

3.处理原始数据

首先从文本中将数据读入，变成我们想要的训练集

def parse_data():
    data = np.loadtxt('data.csv')
    dataMat = data[:, 0:-1]
    classLabels = data[:, -1]
    dataMat = np.insert(dataMat, 0, 1, axis=1)
    return dataMat, classLabels

注意dataMat = np.insert(dataMat, 0, 1, axis=1)这一行的作用是插入偏置项${\omega }_0$。原始数据集中是二维特征，加入偏置项以后变成了三维。

4.定义sigmoid函数与loss function

接下来定义sigmoid 函数与loss function

def sigmoid(x):
    return 1 / (1 + np.exp(-x))

def loss_funtion(dataMat, classLabels, weights):
    m, n = np.shape(dataMat)
    loss = 0.0
    for i in range(m):
        sum_theta_x = 0.0
        for j in range(n):
            sum_theta_x += dataMat[i, j] * weights.T[0, j]
        propability = sigmoid(sum_theta_x)
        loss += -classLabels[i, 0] * np.log(propability) - (1 - classLabels[i, 0]) * np.log(1 - propability)
    return loss

注意loss function的公式，可以从参考文献2中查找详细推导过程。

5.定义梯度下降过程

定义梯度下降过程，并且画出迭代过程loss function的变化曲线。完整代码如下

import numpy as np
import matplotlib.pyplot as plt


def sigmoid(x):
    return 1 / (1 + np.exp(-x))


def parse_data():
    data = np.loadtxt('data.csv')
    dataMat = data[:, 0:-1]
    classLabels = data[:, -1]
    dataMat = np.insert(dataMat, 0, 1, axis=1)
    return dataMat, classLabels


def loss_funtion(dataMat, classLabels, weights):
    m, n = np.shape(dataMat)
    loss = 0.0
    for i in range(m):
        sum_theta_x = 0.0
        for j in range(n):
            sum_theta_x += dataMat[i, j] * weights.T[0, j]
        propability = sigmoid(sum_theta_x)
        loss += -classLabels[i, 0] * np.log(propability) - (1 - classLabels[i, 0]) * np.log(1 - propability)
    return loss


def grad_descent(dataMatIn, classLabels):
    dataMatrix = np.mat(dataMatIn)  #(m,n)
    labelMat = np.mat(classLabels).T
    m, n = np.shape(dataMatrix)
    weights = np.ones((n, 1))
    alpha = 0.01
    maxstep = 10000
    eps = 0.0001
    count = 0
    loss_array = []

    for i in range(maxstep):
        loss = loss_funtion(dataMatrix, labelMat, weights)

        h_theta_x = sigmoid(dataMatrix * weights)
        e = h_theta_x - labelMat
        new_weights = weights - alpha * dataMatrix.T * e
        new_loss = loss_funtion(dataMatrix, labelMat, new_weights)
        loss_array.append(new_loss)
        if abs(new_loss - loss) < eps:
            break
        else:
            weights = new_weights
            count += 1

    print "count is: ", count
    print "loss is: ", loss
    print "weights is: ", weights

    return weights, loss_array

def plotloss(loss_array):
    n = len(loss_array)
    plt.xlabel("iteration num")
    plt.ylabel("loss")
    plt.scatter(range(1, n+1), loss_array)
    plt.show()

data, labels = parse_data()
r, loss_array = grad_descent(data, labels)
r = np.mat(r).transpose()
plotloss(loss_array)

将代码run起来以后，输出如下：

count is:  797
loss is:  9.393884167889501
weights is:  [[13.11270842]
 [ 1.1410689 ]
 [-1.78438114]]

得到的图像如下:

可以看出在迭代过程中，loss function还是比较稳定地变小的。

参考文献
1.https://blog.csdn.net/bitcarmanlee/article/details/51473567
2.https://blog.csdn.net/bitcarmanlee/article/details/51165444