人工智能实战2019_第五次作业_龚楚

标题	内容
这个作业属于哪个课程	人工智能实战2019
这个作业的要求在哪里	第五次作业
我在这个课程的目标是	掌握人工智能方法基本原理
这个作业在哪个具体方面帮助我实现目标	神经网络模型基础：通过训练与门与或门实现简单分类模型(线性二分类)
Reference	神经网络基本原理简明教程

作业正文：

import numpy as np
import matplotlib.pyplot as plt

def ReadData_and():
    X = np.array([1,1,0,0,1,0,1,0]).reshape(2,4)
    Y = np.array([1,0,0,0]).reshape(1,4)
    return X,Y

def ReadData_or():
    X = np.array([1,1,0,0,1,0,1,0]).reshape(2,4)
    Y = np.array([1,1,1,0]).reshape(1,4)
    return X,Y

def Sigmoid(x):
    s = 1/(1+np.exp(-x))
    return s

def ForwardCalculation(W,B,X):
    z = np.dot(W,X) + B
    a = Sigmoid(z)
    return a

def BackPropagation(X,y,A):
    m = X.shape[1]
    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

#cross entropy loss - binary
def CheckLoss(W, B, X, Y):
    count = X.shape[1]
    A = ForwardCalculation(W,B,X)
    p0 = np.multiply(1-Y,np.log(1-A))
    p1 = np.multiply(Y,np.log(A))
    LOSS = np.sum(-(p0 + p1))
    loss = LOSS / count
    return loss

def ShowResult(W,B,X,Y,title):
    # 根据w,b值画出分割线
    w = -W[0,0]/W[0,1]
    b = -B[0,0]/W[0,1]
    x = np.array([0,1])
    y = w * x + b
    plt.plot(x,y)

    # 画出原始样本值
    for i in range(X.shape[1]):
        if Y[0,i] == 0:
            plt.scatter(X[0,i],X[1,i],marker="o",c='b',s=64)
        else:
            plt.scatter(X[0,i],X[1,i],marker="^",c='r',s=64)
    plt.axis([-0.1,1.1,-0.1,1.1])
    plt.title(title)
    plt.show()

if __name__ == '__main__':
    #X, Y = ReadData_and()
    X, Y = ReadData_or()
    num_features = X.shape[0]
    num_example = X.shape[1]
    
    #w, b = np.random.random(),np.random.random()
    W = np.zeros(num_features)
    B = 0.0
    
    eta = 0.1
    eps = 1e-2
    max_epoch = 5000
    loss = 1
    loss_list = []
    
    for epoch in range(max_epoch):
        for i in range(num_example):
            x = X[:,i].reshape(num_features,1)
            y = Y[:,i].reshape(1,1)
            
            a = ForwardCalculation(W,B,x)
            dW, dB = BackPropagation(x,y,a)
            W, B = UpdateWeights(W, B, dW, dB, eta)
        
        loss = CheckLoss(W, B, X, Y)
        loss_list.append(loss)
        if loss < eps:
            break
    
    #cross entropy loss figure
    plt.figure()
    plt.plot(loss_list)
    plt.xlabel('epoch')
    plt.ylabel('cross entropy loss')
    plt.show()
    
    #classification result
    A = ForwardCalculation(W,B,X)
    Class = np.around(A)
    print(A)
    print(Class)
    
    #visualize
    #ShowResult(W,B,X,Y,'AND')
    ShowResult(W,B,X,Y,'OR')

运行结果