Evaluation - Preprocessing

[biochall...@gmail.com]

Hi,

i saw in the document evaluation example, that for the cross validation the preprocessing of the data is repeated every time, when "train()" is called. Are there any concerns, when i preprocess the data only one time and serialize the"raw instances" and for the cross validation i load the the instances  and split them into train and test instances?

Best Regards,