BP神经网络分类算法代码
时间: 2023-09-09 16:13:06 浏览: 41
以下是一个简单的BP神经网络分类算法的代码示例:
```python
import numpy as np
# 定义sigmoid函数
def sigmoid(x):
return 1 / (1 + np.exp(-x))
# 定义BP神经网络类
class BPNeuralNetwork:
def __init__(self, input_nodes, hidden_nodes, output_nodes):
# 初始化权重和偏置
self.weights_ih = np.random.normal(0.0, pow(input_nodes, -0.5), (hidden_nodes, input_nodes))
self.weights_ho = np.random.normal(0.0, pow(hidden_nodes, -0.5), (output_nodes, hidden_nodes))
self.bias_h = np.random.normal(0.0, 1.0, (hidden_nodes, 1))
self.bias_o = np.random.normal(0.0, 1.0, (output_nodes, 1))
# 前向传播
def forward(self, inputs):
inputs = np.array(inputs, ndmin=2).T
hidden_inputs = np.dot(self.weights_ih, inputs) + self.bias_h
hidden_outputs = sigmoid(hidden_inputs)
final_inputs = np.dot(self.weights_ho, hidden_outputs) + self.bias_o
final_outputs = sigmoid(final_inputs)
return final_outputs
# 反向传播
def backward(self, inputs, targets, learning_rate):
inputs = np.array(inputs, ndmin=2).T
targets = np.array(targets, ndmin=2).T
hidden_inputs = np.dot(self.weights_ih, inputs) + self.bias_h
hidden_outputs = sigmoid(hidden_inputs)
final_inputs = np.dot(self.weights_ho, hidden_outputs) + self.bias_o
final_outputs = sigmoid(final_inputs)
output_errors = targets - final_outputs
hidden_errors = np.dot(self.weights_ho.T, output_errors)
self.weights_ho += learning_rate * np.dot((output_errors * final_outputs * (1 - final_outputs)), np.transpose(hidden_outputs))
self.bias_o += learning_rate * output_errors
self.weights_ih += learning_rate * np.dot((hidden_errors * hidden_outputs * (1 - hidden_outputs)), np.transpose(inputs))
self.bias_h += learning_rate * hidden_errors
# 训练模型
def train(self, inputs_list, targets_list, epochs, learning_rate):
for i in range(epochs):
for inputs, targets in zip(inputs_list, targets_list):
self.backward(inputs, targets, learning_rate)
# 预测
def predict(self, inputs_list):
predictions = []
for inputs in inputs_list:
prediction = self.forward(inputs)
predictions.append(np.argmax(prediction))
return predictions
```
使用示例:
```python
# 创建BP神经网络
nn = BPNeuralNetwork(2, 3, 2)
# 训练数据
inputs = np.array([[0, 0], [0, 1], [1, 0], [1, 1]])
targets = np.array([[0, 1], [1, 0], [1, 0], [0, 1]])
# 训练模型
nn.train(inputs, targets, epochs=10000, learning_rate=0.1)
# 预测
predictions = nn.predict(inputs)
print(predictions) # 输出:[0, 1, 1, 0]
```
上述代码实现了一个基本的BP神经网络分类算法,用于解决二分类问题。其中,`inputs`和`targets`是训练数据,`epochs`表示训练轮数,`learning_rate`表示学习率。在训练完成后,可以使用`predict`方法对输入数据进行预测。