人工智能实战第五次作业_李大

项目 内容
课程 人工智能实战2019
作业要求 人工智能实战第五次作业
我的课程目标 第一次作业 介绍自己,提出课程项目建议
此作业帮助 熟悉反向传播,sigmoid激活函数
我的Github主页 LeeDua

代码

import numpy as np
import matplotlib.pyplot as plt

def Sigmiod(x):
    A = 1/(1+np.exp(-x))
    return A

def ForwardCalculation(w,b,x):
    z = np.dot(w,x) + b
    A = Sigmiod(z)
    return z, A

def BackPropagation(x,y,A,m):
    dZ = A - y
    dB = dZ.sum(axis = 1,keepdims = True)/m
    dW = np.dot(dZ, x.T)/m
    return dW, dB

def UpdateWeights(w, b, dW, dB, eta):
    w = w - eta*dW
    b = b - eta*dB
    return w,b

def CheckLoss(A, Y, m):
    Loss = np.sum(-(np.multiply(Y, np.log(A)) + np.multiply((1 - Y), np.log(1 - A))))/m
    return Loss

def ShowResult(X, Y, w, b, m):
    for i in range(m):
        if Y[i] == 0:
            plt.plot(X[0,i], X[1,i], '^', c='r')
        else:
            plt.plot(X[0,i], X[1,i], 'x', c='g')
    x = np.linspace(-0.1,1.1,100)
    y = - (w[0,0] / w[0,1]) * x - (b[0,0] / w[0,1])
    plt.plot(x,y)
    plt.axis([-0.1,1.1,-0.1,1.1])
    plt.show()


def GetSample(GateType):
    X_And = np.array([0, 0, 1, 1, 0, 1, 0, 1]).reshape(2, 4)
    y_And = np.array([0, 0, 0, 1])

    X_Or = np.array([0,0,1,1,0,1,0,1]).reshape(2,4)
    y_Or = np.array([0,1,1,1])
    if GateType == "or":
        return X_Or,y_Or
    elif GateType == "and":
        return X_And,y_And

if __name__ == '__main__':
    eta = 0.1
    eps = 1e-4
    max_epoch = 10000
    loss = 1
    X,Y = GetSample("and")

    num_features = X.shape[0]
    num_example = X.shape[1]
    w = np.zeros((1,num_features))
    b = np.zeros((1,1))
    for epoch in range(max_epoch):
        Z, A = ForwardCalculation(w,b,X)
        dW, dB = BackPropagation(X,Y,A,num_example)
        w, b = UpdateWeights(w, b, dW, dB, eta)
        loss = CheckLoss(A, Y, num_example)
        print(epoch,loss)
        if loss < eps:
            break

    ShowResult(X, Y, w, b, num_example)

结果

  • 与门 iter:9999 Loss:0.017445056023741874

  • 或门 iter:9999 Loss:0.009305581322240441

  • 可以看到10000次迭代后都能很好地达到预期分类效果

posted @ 2019-04-08 11:52  lixiaoda  阅读(152)  评论(0编辑  收藏  举报