Generalization analysis with deep ReLU networks for metric and similarity learning

arXiv:2405.06415v1 Announce Type: cross
Abstract: While considerable theoretical progress has been devoted to the study of metric and similarity learning, the generalization mystery is still missing. In this paper, we study the generalization performance of metric and similarity learning by leveraging the specific structure of the true metric (the target function). Specifically, by deriving the explicit form of the true metric for metric and similarity learning with the hinge loss, we construct a structured deep ReLU neural network as an …

abstract analysis arxiv cs.lg networks paper performance progress relu stat.ml study true type while