$f$-Divergence Based Classification: Beyond the Use of Cross-Entropy

arXiv:2401.01268v2 Announce Type: replace
Abstract: In deep learning, classification tasks are formalized as optimization problems often solved via the minimization of the cross-entropy. However, recent advancements in the design of objective functions allow the usage of the $f$-divergence to generalize the formulation of the optimization problem for classification. We adopt a Bayesian perspective and formulate the classification task as a maximum a posteriori probability problem. We propose a class of objective functions based on the variational representation of the $f$-divergence. …

abstract arxiv bayesian beyond classification cross-entropy cs.lg deep learning design divergence eess.sp entropy functions however optimization replace tasks type usage via