Ziquan OU

CODA: A Correlation-Oriented Disentanglement and Augmentation Modeling Scheme for Better Resisting Subpopulation Shifts

Data-driven models learned often struggle to generalize due to widespread subpopulation shifts, especially the presence of both spurious correlations and group imbalance (SC-GI). To learn models more powerful for defending against SC-GI, we propose a {\bf Correlation-Oriented Disentanglement and Aug…