Abstract | ||
---|---|---|
The developments of Rademacher complexity and PAC-Bayesian theory have been largely independent. One exception is the PAC-Bayes theorem of Kakade, Sridharan, and Tewari [21], which is established via Rademacher complexity theory by viewing Gibbs classifiers as linear operators. The goal of this paper is to extend this bridge between Rademacher complexity and state-of-the-art PAC-Bayesian theory. We first demonstrate that one can match the fast rate of Catoni's PAC-Bayes bounds [8] using shifted Rademacher processes [27, 43, 44]. We then derive a new fast-rate PAC-Bayes bound in terms of the "flatness" of the empirical risk surface on which the posterior concentrates. Our analysis establishes a new framework for deriving fast-rate PAC-Bayes bounds and yields new insights on PAC-Bayesian theory. |
Year | Venue | Keywords |
---|---|---|
2019 | ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 32 (NIPS 2019) | rademacher complexity,linear operators,bayes theorem |
Field | DocType | Volume |
Flatness (systems theory),Discrete mathematics,Mathematical optimization,Computer science,Rademacher complexity,Operator (computer programming),Bayes' theorem | Conference | 32 |
ISSN | Citations | PageRank |
1049-5258 | 0 | 0.34 |
References | Authors | |
0 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jun Yang | 1 | 3 | 1.76 |
Shengyang Sun | 2 | 27 | 4.06 |
Daniel M. Roy | 3 | 818 | 63.27 |