Hello,
Can anyone explain why the default data division for the trainAutoencoder function (example usage here Train Stacked Autoencoders for Image Classification) uses the default data division method "dividetrain", i.e., data division is training only, rather than a training/validation/testing one (e.g., dividerand, divideint, etc.)?
Thanks.Hello,
Can anyone explain why the default data division for the trainAutoencoder function (example usage here Train Stacked Autoencoders for Image Classification) uses the default data division method "dividetrain", i.e., data division is training only, rather than a training/validation/testing one (e.g., dividerand, divideint, etc.)?
Thanks. Hello,
Can anyone explain why the default data division for the trainAutoencoder function (example usage here Train Stacked Autoencoders for Image Classification) uses the default data division method "dividetrain", i.e., data division is training only, rather than a training/validation/testing one (e.g., dividerand, divideint, etc.)?
Thanks. autoencoder, data division MATLAB Answers — New Questions

​