Generative adversarial learning with optimal input dimension and its adaptive generator architecture

arXiv:2405.03723v1 Announce Type: new
Abstract: We investigate the impact of the input dimension on the generalization error in generative adversarial networks (GANs). In particular, we first provide both theoretical and practical evidence to validate the existence of an optimal input dimension (OID) that minimizes the generalization error. Then, to identify the OID, we introduce a novel framework called generalized GANs (G-GANs), which includes existing GANs as a special case. By incorporating the group penalty and the architecture penalty developed in …

abstract adversarial adversarial learning architecture arxiv cs.lg error evidence gans generative generative adversarial networks generator impact networks practical stat.me stat.ml type