Mhammedi, Z; Grunwald, PD; Guedj, B; (2019) PAC-Bayes Un-Expected Bernstein Inequality. In: Proceedings of the Thirty-third Conference on Neural Information Processing Systems 2019. (pp. p. 9387). NIPS: Vancouver, Canada..

Preview

Text 1905.13367v1.pdf - Accepted Version Download (401kB) | Preview

Abstract

We present a new PAC-Bayesian generalization bound. Standard bounds contain a $\sqrt{L_n \cdot \operatorname{KL}/n}$ complexity term which dominates unless $L_n$, the empirical error of the learning algorithm's randomized predictions, vanishes. We manage to replace $L_n$ by a term which vanishes in many more situations, essentially whenever the employed learning algorithm is sufficiently stable on the dataset at hand. Our new bound consistently beats state-of-the-art bounds both on a toy example and on UCI datasets (with large enough $n$). Theoretically, unlike existing bounds, our new bound can be expected to converge to $0$ faster whenever a Bernstein/Tsybakov condition holds, thus connecting PAC-Bayesian generalization and excess risk bounds --- for the latter it has long been known that faster convergence can be obtained under Bernstein conditions. Our main technical tool is a new concentration inequality which is like Bernstein's but with $X^2$ taken outside its expectation.

Download activity - last month

Export as XMLCSVJSON

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item