MMD Aggregated Two-Sample Test - Université Toulouse - Jean Jaurès Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2021

MMD Aggregated Two-Sample Test

Résumé

We propose a novel nonparametric two-sample test based on the Maximum Mean Discrepancy (MMD), which is constructed by aggregating tests with different kernel bandwidths. This aggregation procedure, called MMDAgg, ensures that test power is maximised over the collection of kernels used, without requiring held-out data for kernel selection (which results in a loss of test power), or arbitrary kernel choices such as the median heuristic. We work in the non-asymptotic framework, and prove that our aggregated test is minimax adaptive over Sobolev balls. Our guarantees are not restricted to a specific kernel, but hold for any product of one-dimensional translation invariant characteristic kernels which are absolutely and square integrable. Moreover, our results apply for popular numerical procedures to determine the test threshold, namely permutations and the wild bootstrap. Through numerical experiments on both synthetic and real-world datasets, we demonstrate that MMDAgg outperforms alternative state-of-the-art approaches to MMD kernel adaptation for two-sample testing.
Fichier principal
Vignette du fichier
2110.15073.pdf (6.63 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03408976 , version 1 (29-10-2021)
hal-03408976 , version 2 (29-06-2022)
hal-03408976 , version 3 (21-08-2023)

Licence

Paternité

Identifiants

Citer

Antonin Schrab, Ilmun Kim, Mélisande Albert, Béatrice Laurent, Benjamin Guedj, et al.. MMD Aggregated Two-Sample Test. 2021. ⟨hal-03408976v1⟩
254 Consultations
239 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More