Hi guys, nice week to everyone, I understood your explanations regards to nested cross validation technique. I would like ask to a couple things about. First question, the other techniques presented to us, should be use with with one of the rules presented to us in previous lessons and the outer technique can be kfold for instance, for example ?
The other questions is any book suggestions to support our learning process?
I am not totally sure to understand your question. If you mean that the choice of the cross-validation should also be considered in nested cross-validation, the answer is yes. If you have non-iid data, then you need both cv to be a TimeSeriesSplit. An identical thing should be done for the stratified or group CV. Indeed, you want to be sure to not get into trouble in the inner CV.
You can refer to Bibliography - #3 by Luz_Mary
ESL is more advanced than the course but this is a good reference.