EUVIMEDEUVIMED entdecken
Crossref

Optimal 1-Wasserstein distance for WGANs

Arthur Stéphanovitch, Ugo Tanielian, Benoît Cadre, Nicolas Klutchnikoff, Gérard Biau

Bernoulli · 01.11.2024 · 30 (4) · 2955-2978

DOI: 10.3150/23-bej1701 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Bernoulli, erschienen am 01.11.2024. Verfasst von Arthur Stéphanovitch und 4 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 35 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The mathematical forces at work behind Generative Adversarial Networks raise challenging theoretical issues. Motivated by the important question of characterizing the geometrical properties of the generated distributions, we provide a thorough analysis of Wasserstein GANs (WGANs) in both the finite sample and asymptotic regimes. We study the specific case where the latent space is univariate and derive results valid regardless of the dimension of the output space. We show in particular that for a fixed sample size, the optimal WGANs are closely linked with connected paths minimizing the sum of the squared Euclidean distances between the sample points. We also highlight the fact that WGANs are able to approach (for the 1-Wasserstein distance) the target distribution as the sample size tends to infinity, at a given convergence rate and provided the family of generative Lipschitz functions grows appropriately. We derive in passing new results on optimal transport theory in the semi-discrete setting.

Zitieren

APA

Stéphanovitch, A., Tanielian, U., Cadre, B., Klutchnikoff, N., & Biau, G. (2024). Optimal 1-Wasserstein distance for WGANs. Bernoulli, 30(4), 2955-2978. https://doi.org/10.3150/23-bej1701

Vancouver

Stéphanovitch A, Tanielian U, Cadre B, Klutchnikoff N, Biau G. Optimal 1-Wasserstein distance for WGANs. Bernoulli. 2024;30(4):2955-2978. doi:10.3150/23-bej1701

BibTeX
@article{Stphanovitch2024,
  author = {Stéphanovitch, Arthur and Tanielian, Ugo and Cadre, Benoît and Klutchnikoff, Nicolas and Biau, Gérard},
  title = {Optimal 1-Wasserstein distance for WGANs},
  journal = {Bernoulli},
  year = {2024},
  volume = {30},
  number = {4},
  pages = {2955--2978},
  doi = {10.3150/23-bej1701}
}
Kostenlos starten →