bp神经网络python实例
时间: 2023-08-31 21:09:30 浏览: 27
以下是一个简单的BP神经网络的Python示例:
```
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 BPNetwork(object):
def __init__(self, input_dim, hidden_dim, output_dim):
# 初始化权重和偏置
self.W1 = np.random.randn(input_dim, hidden_dim) / np.sqrt(input_dim)
self.b1 = np.zeros((1, hidden_dim))
self.W2 = np.random.randn(hidden_dim, output_dim) / np.sqrt(hidden_dim)
self.b2 = np.zeros((1, output_dim))
def feedforward(self, X):
# 前向传播
self.z1 = np.dot(X, self.W1) + self.b1
self.a1 = sigmoid(self.z1)
self.z2 = np.dot(self.a1, self.W2) + self.b2
self.y_hat = sigmoid(self.z2)
return self.y_hat
def backpropagation(self, X, y, learning_rate=0.1):
# 反向传播
delta3 = (self.y_hat - y) * sigmoid(self.z2) * (1 - sigmoid(self.z2))
dW2 = np.dot(self.a1.T, delta3)
db2 = np.sum(delta3, axis=0, keepdims=True)
delta2 = np.dot(delta3, self.W2.T) * sigmoid(self.z1) * (1 - sigmoid(self.z1))
dW1 = np.dot(X.T, delta2)
db1 = np.sum(delta2, axis=0)
# 更新权重和偏置
self.W2 += -learning_rate * dW2
self.b2 += -learning_rate * db2
self.W1 += -learning_rate * dW1
self.b1 += -learning_rate * db1
def train(self, X, y, epochs=1000, learning_rate=0.1):
# 训练模型
for i in range(epochs):
self.feedforward(X)
self.backpropagation(X, y, learning_rate)
def predict(self, X):
# 预测
return self.feedforward(X)
# 示例数据
X = np.array([[0, 0], [0, 1], [1, 0], [1, 1]])
y = np.array([[0], [1], [1], [0]])
# 构建BP神经网络
bp = BPNetwork(2, 4, 1)
# 训练模型
bp.train(X, y)
# 预测
print(bp.predict(X))
```
这个示例实现了一个两层的BP神经网络,用于解决异或逻辑运算问题。输入层有两个节点,隐藏层有四个节点,输出层有一个节点。示例数据包括四个输入样本和对应的输出,模型经过训练后可以正确地预测样本的输出。