cnn 全连接层 python

以下是使用 Python 实现 CNN 和全连接层的示例代码：

import numpy as np

# 定义全连接层类
class FullyConnectedLayer:
    def __init__(self, input_size, output_size):
        self.weights = np.random.rand(input_size, output_size)
        self.bias = np.zeros(output_size)
        
    def forward(self, x):
        return np.dot(x, self.weights) + self.bias

# 定义卷积层类
class ConvolutionalLayer:
    def __init__(self, input_size, num_filters, filter_size, stride):
        self.input_size = input_size
        self.num_filters = num_filters
        self.filter_size = filter_size
        self.stride = stride
        self.filters = np.random.rand(num_filters, input_size, filter_size, filter_size)
        
    def forward(self, x):
        batch_size, channels, height, width = x.shape
        output_height = int((height - self.filter_size) / self.stride) + 1
        output_width = int((width - self.filter_size) / self.stride) + 1
        
        # 初始化输出
        output = np.zeros((batch_size, self.num_filters, output_height, output_width))
        
        # 对每个样本进行卷积操作
        for i in range(batch_size):
            for j in range(self.num_filters):
                for h in range(output_height):
                    for w in range(output_width):
                        vertical_start = h * self.stride
                        vertical_end = vertical_start + self.filter_size
                        horizontal_start = w * self.stride
                        horizontal_end = horizontal_start + self.filter_size
                        output[i, j, h, w] = np.sum(x[i, :, vertical_start:vertical_end, horizontal_start:horizontal_end] * self.filters[j])        
        return output

以上代码中，`FullyConnectedLayer` 类表示全连接层，它的前向传播方法 `forward` 接受输入数据 `x`，并输出经过权重和偏置计算后的结果。`ConvolutionalLayer` 类表示卷积层，它的前向传播方法 `forward` 接受输入数据 `x`，并输出经过卷积计算后的结果。

需要注意的是，以上代码仅为示例代码，实际使用时需要考虑更多的细节和实现方式。