Testing the generalization capacity of an algorithm obtained is crucial for any prediction methodology, for which cross-validation methods were developed. However, when dealing with data that have dependency among observations, ...