Communication Dans Un Congrès Année : 2025

Lattice Climber Attack: Adversarial attacks for randomized mixtures of classifiers

Résumé

Finite mixtures of classifiers (a.k.a. randomized ensembles) have been proposed as a way to improve robustness against adversarial attacks. However, existing attacks have been shown to not suit this kind of classifier. In this paper, we discuss the problem of attacking a mixture in a principled way and introduce two desirable properties of attacks based on a geometrical analysis of the problem (effectiveness and maximality). We then show that existing attacks do not meet both of these properties. Finally, we introduce a new attack called lattice climber attack with theoretical guarantees in the binary linear setting, and demonstrate its performance by conducting experiments on synthetic and real datasets.

Fichier principal
Vignette du fichier
main.pdf (2.5 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-05089941 , version 1 (29-05-2025)
hal-05089941 , version 2 (13-06-2025)
hal-05089941 , version 3 (16-01-2026)

Licence

Identifiants

  • HAL Id : hal-05089941 , version 2

Citer

Lucas Gnecco, Benjamin Negrevergne, Yann Chevaleyre. Lattice Climber Attack: Adversarial attacks for randomized mixtures of classifiers. Proceedings of ECML PKDD 2025, Sep 2025, Porto, Portugal. ⟨hal-05089941v2⟩
618 Consultations
407 Téléchargements

Partager

  • More