深度学习之路二 将上一篇的逻辑流水线变成稍微通用的模型

import numpy as np

class NeuralNetwork:
    def __init__(self, input_size, output_size):
        self.input_size = input_size
        self.output_size = output_size
        self.weights = np.random.rand(input_size, output_size)
        self.biases = np.random.rand(1, output_size)

    def forward(self, inputs):
        self.z = np.dot(inputs, self.weights) + self.biases
        self.output = np.tanh(self.z)
        return self.output

    def tanh_derivative(self):
        return 1 - np.power(self.output, 2)
    
    def backward(self, inputs, labels, learning_rate):
        error = self.output - labels
        d_weights = np.dot(inputs.T, error * self.tanh_derivative())
        d_biases = np.sum(error * (1 - np.power(self.output, 2)), axis=0, keepdims=True)
        self.weights -= learning_rate * d_weights
        self.biases -= learning_rate * d_biases

    def train(self, inputs, labels, learning_rate, epochs):
        for i in range(epochs):
            output = self.forward(inputs)
            self.backward(inputs, labels, learning_rate)
            loss = np.mean(np.square(labels - output))
            print("Epoch {}, Loss: {}".format(i, loss))

    def print(self):
        print("Weights: ", self.weights)
        print("Biases: ", self.biases)
        print("Output: ", self.output)

# Example usage
inputs = np.array([[1.0, 2.0]])
labels = np.array([[1.0]])
network = NeuralNetwork(input_size=2, output_size=1)
network.train(inputs, labels, learning_rate=0.1, epochs=100)
network.print()