MATLAB Answers

How to make loose function for validation and training togther for autoencoder

조회 수: 3(최근 30일)
Abdussalam Elhanashi
Abdussalam Elhanashi 2020년 11월 9일
Hi
I am working autoencoder and I am looking to make loose function curve for training and validation together while training the model
I checked this Example of training autoencoder but only show loose curve for training
https://www.mathworks.com/help/deeplearning/ref/trainautoencoder.html
How to make Training and Validation loose curve togther while training the model
Herein my code
dataDir= fullfile('Data/');
exts = {'.jpg','.png','.tif','BMP'};
imds = imageDatastore(fullfile(dataDir),...
'IncludeSubfolders',true,'FileExtensions','.jpg','LabelSource','foldernames');
countEachLabel(imds);
[TrainData, TestData, ValidData] = splitEachLabel(imds, 0.10, 0.20);
size(TrainData);
countEachLabel(TrainData);
numImages = numel(TrainData.Files);
for i = 1:numImages
img = readimage(TrainData, i);
img=rgb2gray(img);
img3= im2double(img);
scale = 0.5;
img8 = imresize(img3,scale);
Train{idx} = (img8); %#ok<SAGROW>
end
numImages = numel(ValidData.Files);
for i = 1:numImages
img = readimage(ValidData, i);
img=rgb2gray(img);
img3= im2double(img);
scale = 0.5;
img8 = imresize(img3,scale);
Valid{idx} = (img8); %#ok<SAGROW>
end
hiddenSize = 50;
autoenc = trainAutoencoder(Valid,Train,hiddenSize,'MaxEpochs',100,...
'DecoderTransferFunction','purelin','EncoderTransferFunction','satlin','L2WeightRegularization',0.004,'SparsityRegularization',4,'SparsityProportion',0.15);

답변(0개)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by