K fold cross validation partition, how to take a set of data and split it up into a training set and a validation set [1]