Construction and evaluation of a classifiers accuracy on a dataset require partitioning labeled data into a training set and a test set. Explain three main methods used for such partitioning.