Fig. 3
From: Learning stable and predictive network-based patterns of schizophrenia and its clinical symptoms
Classification of patients vs. controls based on ss-link-weight features. Y-axis shows the average CV error over 95 leave-subject-out folds (see the Supplementary Fig. S4 for the corresponding FP and FN rates) for different classifiers, and for varying size of top-ranking feature subset shown on x-axis
