Fig. 3From: Machine learning analysis of motor evoked potential time series to predict disability progression in multiple sclerosisPipeline Schematic overview of the data analysis pipeline. By grouped stratified shufflesplit we mean splits of train/test sets generated by assigning samples to train or test set at random, but keeping the patients in test and training set separated, and making sure the ratio of positive targets are roughly the same for the test and training set. To assess the impact of the size of the training set, we run this pipeline for 4 different ratios of train/test set sizeBack to article page