输出基于python的一层隐藏层的bp神经网络代码

以下是基于 Python 的一层隐藏层 BP 神经网络代码示例：

import numpy as np

# 定义一层隐藏层的 BP 神经网络类
class NeuralNetwork:
    def __init__(self, input_size, hidden_size, output_size):
        self.input_size = input_size
        self.hidden_size = hidden_size
        self.output_size = output_size
        
        # 初始化权重和偏置
        self.weights_input_to_hidden = np.random.normal(scale=1/input_size**0.5, size=(input_size, hidden_size))
        self.weights_hidden_to_output = np.random.normal(scale=1/hidden_size**0.5, size=(hidden_size, output_size))
        self.bias_hidden = np.zeros(hidden_size)
        self.bias_output = np.zeros(output_size)

    # 定义 sigmoid 函数
    def sigmoid(self, x):
        return 1 / (1 + np.exp(-x))

    # 实现前向传播
    def forward(self, inputs):
        hidden_inputs = np.dot(inputs, self.weights_input_to_hidden) + self.bias_hidden  # 隐藏层输入
        hidden_outputs = self.sigmoid(hidden_inputs)  # 隐藏层输出
        final_inputs = np.dot(hidden_outputs, self.weights_hidden_to_output) + self.bias_output  # 输出层输入
        final_outputs = final_inputs  # 输出层输出
        return final_outputs
    
    # 实现反向传播
    def backward(self, inputs, targets, outputs, learning_rate):
        error = targets - outputs  # 计算误差
        
        output_error_term = error  # 输出层误差项
        hidden_error = np.dot(self.weights_hidden_to_output, output_error_term)  # 隐藏层误差
        hidden_error_term = hidden_error * hidden_outputs * (1 - hidden_outputs)  # 隐藏层误差项
        
        delta_weights_hidden_to_output = learning_rate * output_error_term * hidden_outputs[:, None]  # 更新隐藏层到输出层的权重
        delta_weights_input_to_hidden = learning_rate * hidden_error_term * inputs[:, None]  # 更新输入层到隐藏层的权重
        
        self.weights_hidden_to_output += delta_weights_hidden_to_output
        self.weights_input_to_hidden += delta_weights_input_to_hidden
        self.bias_output += learning_rate * output_error_term  # 更新输出层偏置项
        self.bias_hidden += learning_rate * hidden_error_term  # 更新隐藏层偏置项

希望以上代码能够满足您的需求。如果还有其他问题，请继续提问。