posted on 2017-09-25, 17:33authored byOscar Esteban, Daniel Birman, Marie Schaer, Oluwasanmi O. Koyejo, Russell A. Poldrack, Krzysztof J. Gorgolewski
The cross-validated area under the curve (AUC) and accuracy (ACC) scores calculated on the ABIDE dataset (train set) are less biased when LoSo is used to create the outer folds, as compares to the evaluation scores obtained in DS030 (held-out set).