Fig. 1
From: Gaussian noise up-sampling is better suited than SMOTE and ADASYN for clinical decision making
The workflow of the study. Models are evaluated with 1000 times repeated MCCV. Three augmentation techniques, namely GNUS, SMOTE, and ADASYN are compared against each other and to the null model. As machine learning models we used logistic regression (LR), random forests (RF), and support-vector machines (SVM) with three different kernels (linear, rbf, and polynomial). As for performance metrics, we used AUC, PR, F1, and MCC