predict participants status to either continued AUD group or remitted AUD group. A supervised linear-kernel SVM that included parameter optimization was trained with a tenfold CV procedure to classify participants into the two groups. The tenfold CV procedure involved randomly dividing the participants into ten equal groups, training the classifier on nine of them, and tested the trained model on the left out one. To ensure randomization of the participants in the calculated model, the dataset was shuffled before every fold. To take advantage of the randomization procedure, we repeated this process ten times, averaging the output results. CV was applied to all models with additional training/testing (70:30) validation analysis to confirm results in the larger samples (EA male and females). Model performance was evaluated by calculating the number of true positives (TP, number of correctly classified remitted AUD) and true negatives (TN, number of correctly classified continued AUD) scores. We computed the classification accuracy as the ratio of sum of TP and TN divided by the sum of all classified subjects. Area under curve (AUC)7 was used to evaluate the classification models. More description of AUC calculation and comparison can be found in Supplementary Materials.
