写一个BP神经网络用python实现的代码

Sure, here's a simple implementation of a BP neural network using Python:

import numpy as np

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
        
        self.weights1 = np.random.rand(self.input_size, self.hidden_size)
        self.bias1 = np.random.rand(self.hidden_size)
        self.weights2 = np.random.rand(self.hidden_size, self.output_size)
        self.bias2 = np.random.rand(self.output_size)
        
    def sigmoid(self, x):
        return 1 / (1 + np.exp(-x))
    
    def feedforward(self, inputs):
        hidden_layer = self.sigmoid(np.dot(inputs, self.weights1) + self.bias1)
        output_layer = self.sigmoid(np.dot(hidden_layer, self.weights2) + self.bias2)
        return output_layer
    
    def train(self, inputs, targets, learning_rate):
        hidden_layer = self.sigmoid(np.dot(inputs, self.weights1) + self.bias1)
        output_layer = self.sigmoid(np.dot(hidden_layer, self.weights2) + self.bias2)
        
        output_error = targets - output_layer
        output_delta = output_error * output_layer * (1 - output_layer)
        
        hidden_error = np.dot(output_delta, self.weights2.T)
        hidden_delta = hidden_error * hidden_layer * (1 - hidden_layer)
        
        self.weights2 += learning_rate * np.dot(hidden_layer.T, output_delta)
        self.bias2 += learning_rate * np.sum(output_delta, axis=0)
        self.weights1 += learning_rate * np.dot(inputs.T, hidden_delta)
        self.bias1 += learning_rate * np.sum(hidden_delta, axis=0)

This implementation uses the sigmoid activation function and backpropagation algorithm to train the neural network. You can initialize the neural network with the number of input, hidden, and output neurons, and then train it with the `train` method using the inputs and target outputs. The `feedforward` method can then be used to predict the output for new inputs.