Part of Advances in Neural Information Processing Systems 36 (NeurIPS 2023) Main Conference Track
Ravid Shwartz-Ziv, Micah Goldblum, Yucen Li, C. Bayan Bruss, Andrew G. Wilson
Real-world datasets are often highly class-imbalanced, which can adversely impact the performance of deep learning models. The majority of research on training neural networks under class imbalance has focused on specialized loss functions and sampling techniques. Notably, we demonstrate that simply tuning existing components of standard deep learning pipelines, such as the batch size, data augmentation, architecture size, pre-training, optimizer, and label smoothing, can achieve state-of-the-art performance without any specialized loss functions or samplers. We also provide key prescriptions and considerations for training under class imbalance, and an understanding of why imbalance methods succeed or fail.