Addressing imbalance in health data: Synthetic minority oversampling using deep learning

Alex X. Wang, Viet-Tuan Le, Hau Nguyen Trung, Binh P. Nguyen

Abstract: Highlights•Class imbalance is a common issue in health data.•We propose a new method called ACVAE to balance data using deep learning.•We also propose an ensemble method using ACVAE and edited undersampling.•The methods improve the accuracy of machine learning models in healthcare.