手把手教你用matlab做深度学习(二)- --CNN

在上一篇博客中，讲解了怎么用matlab搭建CNN网络模型，并给出了训练过程与结果。但是结果不是很满意，大概训练精度在80%左右，现在给出改进方案。

1.首先，我们可以把CNN滤波输出数改大点，从原来的32改为numFilters = 128，完整的程序如下：

[trainingImages,trainingLabels,testImages,testLabels] = helperCIFAR10Data.load('cifar10Data');
numImageCategories = 10;
categories(trainingLabels)
[height, width, numChannels, ~] = size(trainingImages);

imageSize = [height width numChannels];
inputLayer = imageInputLayer(imageSize)
% Convolutional layer parameters
filterSize = [5 5];
numFilters = 128;

middleLayers = [
    
% The first convolutional layer has a bank of 32 5x5x3 filters. A
% symmetric padding of 2 pixels is added to ensure that image borders
% are included in the processing. This is important to avoid
% information at the borders being washed away too early in the
% network.
convolution2dLayer(filterSize, numFilters, 'Padding', 2)

% Note that the third dimension of the filter can be omitted because it
% is automatically deduced based on the connectivity of the network. In
% this case because this layer follows the image layer, the third
% dimension must be 3 to match the number of channels in the input
% image.

% Next add the ReLU layer:
reluLayer()

% Follow it with a max pooling layer that has a 3x3 spatial pooling area
% and a stride of 2 pixels. This down-samples the data dimensions from
% 32x32 to 15x15.
maxPooling2dLayer(3, 'Stride', 2)

% Repeat the 3 core layers to complete the middle of the network.
convolution2dLayer(filterSize, numFilters, 'Padding', 2)
reluLayer()
maxPooling2dLayer(3, 'Stride',2)

convolution2dLayer(filterSize, 2 * numFilters, 'Padding', 2)
reluLayer()
maxPooling2dLayer(3, 'Stride',2)

]

finalLayers = [
    
% Add a fully connected layer with 64 output neurons. The output size of
% this layer will be an array with a length of 64.
fullyConnectedLayer(256)

% Add an ReLU non-linearity.
reluLayer

% Add the last fully connected layer. At this point, the network must
% produce 10 signals that can be used to measure whether the input image
% belongs to one category or another. This measurement is made using the
% subsequent loss layers.
fullyConnectedLayer(numImageCategories)

% Add the softmax loss layer and classification layer. The final layers use
% the output of the fully connected layer to compute the categorical
% probability distribution over the image classes. During the training
% process, all the network weights are tuned to minimize the loss over this
% categorical distribution.
softmaxLayer
classificationLayer
]

layers = [
    inputLayer
    middleLayers
    finalLayers
    ]

layers(2).Weights = 0.0001 * randn([filterSize numChannels numFilters]);

% Set the network training options
opts = trainingOptions('sgdm', ...
    'Momentum', 0.9, ...
    'InitialLearnRate', 0.001, ...
    'LearnRateSchedule', 'piecewise', ...
    'LearnRateDropFactor', 0.1, ...
    'LearnRateDropPeriod', 8, ...
    'L2Regularization', 0.004, ...
    'MaxEpochs', 40, ...
    'MiniBatchSize', 128, ...
    'Verbose', true,...
 'Plots','training-progress');

% A trained network is loaded from disk to save time when running the
% example. Set this flag to true to train the network.
doTraining = true;

if doTraining    
    % Train a network.
    cifar10Net = trainNetwork(trainingImages, trainingLabels, layers, opts);
else
    % Load pre-trained detector for the example.
    load('rcnnStopSigns.mat','cifar10Net')       
end

% Extract the first convolutional layer weights
w = cifar10Net.Layers(2).Weights;

% rescale the weights to the range [0, 1] for better visualization
w = rescale(w);

figure
montage(w)

% Run the network on the test set.
YTest = classify(cifar10Net, testImages);

% Calculate the accuracy.
accuracy = sum(YTest == testLabels)/numel(testLabels)

训练结果如下：

从图中可以看出结果非常好，训练结果达到96%以上，然而测试结果基本没有提高，

 

 

现在这种情况属于High variance,这是由于过拟合造成的，现在要去解决这个问题。