Reviews: Feature selection in functional data classification with recursive maxima hunting

First let me say that while I know feature selection well, I haven't had the opportunity to read a lot of papers about functional feature selection. Therefore my background here is limited in terms of references. The paper is well organized and clear, in my opinion even though ironically wrt the contents it has a few redundancies - but that's fine. The experiments seem to be carried out thoroughly with one exception as the authors note themselves: the parameters s and r seem to be chosen a little arbitrarily and in my opinion more values of those should have been included in the cross validation procedure. It is also not very clear whether these values are fixed or not during the experiments on real data.