Is there any way to add cross validation in trainingOptions function while using DNN?

Currently, I am training a CNN model to classify images. I am using splitEachLabel function to split the dataset into two segments. Training, validation. Then using augmentedImageDatastore for each set. Lastly using trainingOptions for setting the parameter and trainNetwork for training the model. Currently, the amount for validation is fixed in the dataset in every epoch (the fixed set of validation data is used). From my knowledge, this is called holdout validation approach.

I am wondering if it is possible to use k-fold corss validation approach rather than holdout validation while training a in deep neural network. If it's yes then how can I do it. How will I apply the dataset into k-fold?

TIA

Answer 1

[Mohammad Sami]

There is no option for cross validation in training options for DNN.

[Mohammad Sami]

With larger data set it may not be worth having k fold cross validation.

https://www.quora.com/Is-cross-validation-heavily-used-in-deep-learning-or-is-it-too-expensive-to-be-used