increase number of hidden layers

Question

Sajitha K.N. 2020-4-27

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/521072-increase-number-of-hidden-layers

回答： Srivardhan Gadila 2020-5-3

This is my matlab code

layers = [

imageInputLayer([256 256 3])

convolution2dLayer(4,13,'Padding','same')

batchNormalizationLayer

reluLayer

maxPooling2dLayer(11,'Stride',11)

convolution2dLayer(7,28,'Padding','same')

batchNormalizationLayer

reluLayer

maxPooling2dLayer(11,'Stride',11)

convolution2dLayer(5,34,'Padding','same')

batchNormalizationLayer

reluLayer

fullyConnectedLayer(38)

softmaxLayer

classificationLayer];

options = trainingOptions('sgdm', ...

'InitialLearnRate',0.06, ...

'MaxEpochs',10, ...

'Shuffle','every-epoch', ...

'ValidationData',imdsValidation, ...

'ValidationFrequency',30, ...

'Verbose',false, ...

'Plots','training-progress');

net = trainNetwork(imdsTrain,layers,options);

YPred = classify(net,imdsValidation);

YValidation = imdsValidation.Labels;

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

I want to increase the no of hidden layers. please somebody help.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Srivardhan Gadila 2020-5-3

2
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/521072-increase-number-of-hidden-layers#answer_429927

在 MATLAB Online 中打开

You can refer to List of Deep Learning Layers and add layers like convolution2dLayer, fullyConnectedLayer to your above layer array.

You can also refer to the following resources Deep Learning Tips and Tricks, Specify Layers of Convolutional Neural Network & Create Simple Deep Learning Network for Classification

layers = [
    imageInputLayer([256 256 3])
    
    convolution2dLayer(4,13,'Padding','same')
    batchNormalizationLayer
    reluLayer
    maxPooling2dLayer(11,'Stride',11)
    
    convolution2dLayer(7,28,'Padding','same')
    batchNormalizationLayer
    reluLayer
    
    maxPooling2dLayer(11,'Stride',11)
    
    convolution2dLayer(5,34,'Padding','same')
    batchNormalizationLayer
    reluLayer
     
    fullyConnectedLayer(38)
    softmaxLayer
    classificationLayer];