用matlab写一个cnn，二分类网络

好的，下面是一个简单的二分类CNN网络的MATLAB实现：

1. 数据准备

首先，我们需要准备训练和测试数据。这里我们使用MATLAB自带的digitDatasetMNIST数据集。读取数据集，并将其分成训练集和测试集：

digitDatasetPath = fullfile(matlabroot,'toolbox','nnet','nndemos', ...
    'nndatasets','DigitDataset');
digitData = imageDatastore(digitDatasetPath, ...
    'IncludeSubfolders',true,'LabelSource','foldernames');

% 将数据集分为训练集和测试集
[trainDigitData,testDigitData] = splitEachLabel(digitData,0.7,'randomized');

2. 构建CNN模型

接着，我们构建CNN模型。这里我们使用了三个卷积层和两个全连接层：

layers = [
    imageInputLayer([28 28 1])
    
    convolution2dLayer(5,20)
    batchNormalizationLayer
    reluLayer
    
    maxPooling2dLayer(2,'Stride',2)
    
    convolution2dLayer(5,50)
    batchNormalizationLayer
    reluLayer
    
    maxPooling2dLayer(2,'Stride',2)
    
    convolution2dLayer(3,100)
    batchNormalizationLayer
    reluLayer
    
    fullyConnectedLayer(500)
    reluLayer
    
    fullyConnectedLayer(2)
    softmaxLayer
    classificationLayer
];

3. 训练模型

接着，我们训练CNN模型。这里我们使用了交叉熵损失函数和Adam优化器：

options = trainingOptions('adam', ...
    'MaxEpochs',10, ...
    'MiniBatchSize',128, ...
    'Shuffle','every-epoch', ...
    'Verbose',false, ...
    'Plots','training-progress');

net = trainNetwork(trainDigitData,layers,options);

4. 测试模型

最后，我们使用测试集来测试CNN模型的性能：

YPred = classify(net,testDigitData);
YTest = testDigitData.Labels;

accuracy = sum(YPred == YTest)/numel(YTest)

完整代码如下：

digitDatasetPath = fullfile(matlabroot,'toolbox','nnet','nndemos', ...
    'nndatasets','DigitDataset');
digitData = imageDatastore(digitDatasetPath, ...
    'IncludeSubfolders',true,'LabelSource','foldernames');

% 将数据集分为训练集和测试集
[trainDigitData,testDigitData] = splitEachLabel(digitData,0.7,'randomized');

layers = [
    imageInputLayer([28 28 1])
    
    convolution2dLayer(5,20)
    batchNormalizationLayer
    reluLayer
    
    maxPooling2dLayer(2,'Stride',2)
    
    convolution2dLayer(5,50)
    batchNormalizationLayer
    reluLayer
    
    maxPooling2dLayer(2,'Stride',2)
    
    convolution2dLayer(3,100)
    batchNormalizationLayer
    reluLayer
    
    fullyConnectedLayer(500)
    reluLayer
    
    fullyConnectedLayer(2)
    softmaxLayer
    classificationLayer
];

options = trainingOptions('adam', ...
    'MaxEpochs',10, ...
    'MiniBatchSize',128, ...
    'Shuffle','every-epoch', ...
    'Verbose',false, ...
    'Plots','training-progress');

net = trainNetwork(trainDigitData,layers,options);

YPred = classify(net,testDigitData);
YTest = testDigitData.Labels;

accuracy = sum(YPred == YTest)/numel(YTest)

希望对你有所帮助！