Fig. 5From: Creating sparser prediction models of treatment outcome in depression: a proof-of-concept study using simultaneous feature selection and hyperparameter tuningPermutation importance from 25 permutations for the most informative clinical features, grouped by classifier and models with and without RFE. Only features that were selected by all 6 clinical models and showed a positive mean importance score (averaged over all 6 models) are presented. The scores show the average decrease in model performance on the test data when a feature was randomly permuted. Error bars represent 95% confidence intervals. LR, logistic regression; RF, random forest classifier; RFE, recursive feature elimination; SVC, support vector classifierBack to article page