I am evaluating the performance of my deep learning model. My model is using an early stopping technic and I would like to do 5-fold cross-validation. Would below be the correct performance evaluation?

training, test split for 4:1 and split the training set into a training, validation set 4:1. Eventually, the train, validation, and test splits are 16:4:5

Do above for 5 times with changing the test set fold.

[ fold 0 ][ fold 1 ][ fold 2 ][ fold 3 ][ fold 4 ]
[                train                 ][  test  ]

# sub folds
[sub  0][sub  1][sub  2][sub  3][sub  4]           <-- split training set into training, validation sets
[            train             ][valid ][  test  ]

I had checked link_01, link_02, link_03, and link_04 but couldn't find exact solution,


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.