Backpropagation python代码详解

Backpropagation

Backpropagation is a method used in artificial neural networks to calculate a gradient that is needed in the calculation of the weights to be used in the network. It is commonly used to train deep neural networks, a term referring to neural networks with more than one hidden layer.

import math
import random
import string
import matplotlib.pyplot as plt
import numpy as np 

random.seed(0)

rand(a, b)保证随机数在[a,b]范围内

def rand(a, b):
    return (b-a)*random.random() + a
print(rand(6, 10))

9.377687406100193

生成I*J的矩阵。

def makeMatrix(I, J, fill=0.0):
    m = []
    for i in range(I):
        m.append([fill]*J)
    return m
m=makeMatrix(3, 5, 4.87)
print(m)

[[4.87, 4.87, 4.87, 4.87, 4.87], [4.87, 4.87, 4.87, 4.87, 4.87], [4.87, 4.87, 4.87, 4.87, 4.87]]

定义激活函数

def sigmoid(x):
    return math.tanh(x)

def dsigmoid(y):
    return 1.0 - y**2

画出来瞧瞧

numpy.arange() 参数的含义：起点，终点，间隔。

numpy.linspace()：起点，终点，以及点个数。

Xaxis = np.arange(-10., 10., 0.2)
Yaxis = []
Zaxis = []
for i in Xaxis:
    Yaxis.append(sigmoid(i))
    Zaxis.append(dsigmoid(i))
plt.plot(Xaxis, Yaxis) 

[<matplotlib.lines.Line2D at 0x2f64a12f278>]

plt.plot(Xaxis, Zaxis) 

[<matplotlib.lines.Line2D at 0x2f64a1d1a20>]

创建NN

一开始的神经网络权重为随机数

class neuralNetwork:
    """三层BP网络"""
    def __init__(self, ni, nh, no):
        # 输入层 隐藏层 输出层
        self.ni = ni + 1
        self.nh = nh
        self.no = no

        # 激活神经网络的所有节点（向量）
        self.ai = [1.0]*self.ni
        self.ah = [1.0]*self.nh
        self.ao = [1.0]*self.no

        # 权重矩阵
        self.wi = makeMatrix(self.ni, self.nh)
        self.wo = makeMatrix(self.nh, self.no)

        # 设置随机值
        for i in range(self.ni):
            for j in range(self.nh):
                self.wi[i][j] = rand(-0.2, 0.2)

        for j in range(self.nh):
            for k in range(self.no):
                self.wo[j][k] = rand(-2.0, 2.0)

        # 最后建立动量因子（矩阵）
        self.ci = makeMatrix(self.ni, self.nh)
        self.co = makeMatrix(self.nh, self.no)
    def weights(self):
        print('输入层权重:')
        for i in range(self.ni):
            print(self.wi[i])
        print()
        print('输出层权重:')
        for j in range(self.nh):
            print(self.wo[j])
    def nodes(self):
        print('各层')
        print(self.ai)
        print(self.ah)
        print(self.ao)

n = neuralNetwork(2, 2, 1)
n.weights()

输入层权重:
[0.10318176117612099, -0.031771367667662004]
[-0.09643329988281467, 0.004509888547443414]
[-0.03802634501983429, 0.11351943561390904]

输出层权重:
[-0.7867490956842902]
[-0.09361218339057675]

有输入数据进行训练可以得到不一样的权重

def update(self, inputs):
    if len(inputs) != self.ni - 1:
        raise  ValueError('与输入层节点数不符！')

    # 激活输入层
    for i in range(self.ni-1):
        self.ai[i] = inputs[i]

    # 激活隐藏层
    for j in range(self.nh):
        sum = 0.0
        for i in range(self.ni):
            sum = sum + self.ai[i] * self.wi[i][j]
        self.ah[j] = sigmoid(sum)

    # 激活输出层
    for k in range(self.no):
        sum = 0.0
        for j in range(self.nh):
            sum = sum + self.ah[j] * self.wo[j][k]
        self.ao[k] = sigmoid(sum)

    return self.ao[:]
neuralNetwork.update = update

data = [
        [0, 0],
        [0, 1],
        [1, 0],
        [1, 1]
    ]
for x in data:
    print(