2021-09-10机器学习感知器

import numpy as np

def andActivator(x):
    if x > 0:
        return 1
    else:
        return 0

class ganzhiqi(object):
    def __init__(self, x):
        self.x = x
        self.w = np.random.rand(2, 1)
        self.b = np.random.rand(1, 1)

    def forward(self):
        self.y = np.zeros((4, 1))
        for i in range(self.x.shape[0]):
            self.y[i] = andActivator(np.sum(self.x[i] * self.w + self.b))

    def backward(self, t, lr):
        delta = np.zeros((4, 1))

        loss = 0

        for i in range(self.x.shape[0]):
            delta[i] = (self.y)[i] - t[i]
            '''
            a=delta[i]
            b=self.y[i]
            c=t[i]
            a=b-c
            '''
            loss -= delta[i]
            gradient_b = delta[i]
            self.b -= lr * gradient_b
            gradient_w = self.x[i].T * gradient_b
            self.w -= lr * gradient_w
        print(loss)
        print(self.w)
        print(self.b)

if __name__ == '__main__':
    x = np.matrix([[1, 1], [1, 0], [0, 1]])
    t = np.matrix([[1], [0], [0],])
    g = ganzhiqi(x)
    lr = 0.01

    for i in range(1000):
        f = g.forward()
        b = g.backward(t, lr)