PAC-Bayesian Generalization Bounds for Adversarial Generative Models

Sokhna Diarra Mbacke; Florence Clerc; Pascal Germain

ICML 2023poster11 citations

PAC-Bayesian Generalization Bounds for Adversarial Generative Models

Sokhna Diarra Mbacke, Florence Clerc, Pascal Germain

Abstract

We extend PAC-Bayesian theory to generative models and develop generalization bounds for models based on the Wasserstein distance and the total variation distance. Our first result on the Wasserstein distance assumes the instance space is bounded, while our second result takes advantage of dimensionality reduction. Our results naturally apply to Wasserstein GANs and Energy-Based GANs, and our bounds provide new training objectives for these two. Although our work is mainly theoretical, we perform numerical experiments showing non-vacuous generalization bounds for Wasserstein GANs on synthetic datasets.

BibTeX

@inproceedings{icml2023_pacbayesiangener,
  title = {PAC-Bayesian Generalization Bounds for Adversarial Generative Models},
  author = {Sokhna Diarra Mbacke and Florence Clerc and Pascal Germain},
  booktitle = {ICML 2023},
  year = {2023}
}