Fig. 1From: Predicting two-year survival versus non-survival after first myocardial infarction using machine learning and Swedish national register dataTraining results. Top panel: Model training result as a function of increasing sample size (1–100%). Bottom panel: Model training performance on the three predictor sets using 100% of training samples (n = 31,166) with the 5 and 10 predictor sets as chosen by each model. Points are mean values of each model’s resampled training runs optimized on the Area Under the Receiver Operating Characteristic (AUROC). Error bars indicate ± SD. C5.0, Boosted C5.0; LR, Logistic regression; RF, Random Forest; SVM, Support Vector MachineBack to article page