I am trying to select "the best" number of WFO cycles. However, the more WFO cycles - the longer the total OOS Test period; and so performance stats are not directly comparable.

What would be the way to achieve common test period across such trials?

Last edited by Zheka; 05/01/22 15:42.