前馈神经网络python实现

下面是一个简单的前馈神经网络的Python实现示例：

import numpy as np

class NeuralNetwork:
    def __init__(self, input_size, hidden_size, output_size):
        self.weights1 = np.random.randn(input_size, hidden_size)
        self.bias1 = np.zeros((1, hidden_size))
        self.weights2 = np.random.randn(hidden_size, output_size)
        self.bias2 = np.zeros((1, output_size))
    
    def forward(self, X):
        self.z1 = np.dot(X, self.weights1) + self.bias1
        self.a1 = np.tanh(self.z1)
        self.z2 = np.dot(self.a1, self.weights2) + self.bias2
        exp_scores = np.exp(self.z2)
        self.probs = exp_scores / np.sum(exp_scores, axis=1, keepdims=True)
        return self.probs
    
    def compute_loss(self, X, y):
        num_examples = len(X)
        log_probs = -np.log(self.probs[range(num_examples), y])
        data_loss = np.sum(log_probs)
        return 1./num_examples * data_loss
    
    def backpropagation(self, X, y, learning_rate=0.01):
        num_examples = len(X)
        delta3 = self.probs
        delta3[range(num_examples), y] -= 1
        dW2 = np.dot(self.a1.T, delta3)
        db2 = np.sum(delta3, axis=0, keepdims=True)
        delta2 = np.dot(delta3, self.weights2.T) * (1 - np.power(self.a1, 2))
        dW1 = np.dot(X.T, delta2)
        db1 = np.sum(delta2, axis=0)
        self.weights1 -= learning_rate * dW1
        self.bias1 -= learning_rate * db1
        self.weights2 -= learning_rate * dW2
        self.bias2 -= learning_rate * db2
    
    def train(self, X, y, num_passes=20000, learning_rate=0.01):
        for i in range(num_passes):
            self.forward(X)
            self.backpropagation(X, y, learning_rate)
            if i % 1000 == 0:
                print("Iteration:", i, " Loss:", self.compute_loss(X, y))

这个前馈神经网络由一个输入层、一个隐藏层和一个输出层组成，其中隐藏层使用tanh作为激活函数，输出层使用softmax作为激活函数。在训练过程中，使用交叉熵损失函数，并使用反向传播算法更新网络参数。