Improved Vapnik Cervonenkis bounds

Olivier Catoni (1)

1 LPMA - Laboratoire de Probabilités et Modèles Aléatoires

We give a new proof of VC bounds where we avoid the use of symmetrization and use a shadow sample of arbitrary size. We also improve on the variance term. This results in better constants, as shown on numerical examples. Moreover our bounds still hold for non identically distributed independent random variables. Keywords: Statistical learning theory, PAC-Bayesian theorems, VC dimension.

Mots clés: Statistical learning theory; PAC-Bayesian theorems; VC dimension

Domaines: Probabilités [math.PR]

Olivier Catoni. Improved Vapnik Cervonenkis bounds. 2004. ⟨hal-00003056⟩