Our recent PNAS paper introduced the Layer-Peeled Model in deep learning that can: 1) predict a hitherto unknown phenomenon termed "Minority Collapse" in imbalanced training; and 2) explain neural collapse discovered by Papyan, Han and Donoho (PNAS, 2020). When minority collapse occurs, the deep learning classifiers on minor classes are simply identical to each other. This limitation can lead to severe fairness issues for minority groups in high-stakes decisionmaking.
About Us
To understand health and disease today, we need new thinking and novel science —the kind we create when multiple disciplines work together from the ground up. That is why this department has put forward a bold vision in population-health science: a single academic home for biostatistics, epidemiology and informatics.
© 2023 Trustees of the University of Pennsylvania. All rights reserved.. | Disclaimer