Random-Walk Perturbations for Online Combinatorial Optimization

Luc Devroye 1 Gábor Lugosi 2, 3 Gergely Neu 4
4 SEQUEL - Sequential Learning
Inria Lille - Nord Europe, CRIStAL - Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL) - UMR 9189
Abstract : We study online combinatorial optimization problems where a learner is interested in minimizing its cumulative regret in the presence of switching costs. To solve such problems, we propose a version of the follow-the-perturbed-leader algorithm in which the cumulative losses are perturbed by independent symmetric random walks. In the general setting, our forecaster is shown to enjoy near-optimal guarantees on both quantities of interest, making it the best known efficient algorithm for the studied problem. In the special case of prediction with expert advice, we show that the forecaster achieves an expected regret of the optimal order O(√ n log N) where n is the time horizon and N is the number of experts, while guaranteeing that the predictions are switched at most O(√ n log N) times, in expectation.
Document type :
Journal articles
Complete list of metadatas

Cited literature [29 references]  Display  Hide  Download

https://hal.inria.fr/hal-01214987
Contributor : Gergely Neu <>
Submitted on : Tuesday, October 13, 2015 - 2:23:15 PM
Last modification on : Friday, March 22, 2019 - 1:35:34 AM
Long-term archiving on : Thursday, April 27, 2017 - 12:16:27 AM

File

fpl_journal_final.pdf
Files produced by the author(s)

Identifiers

Citation

Luc Devroye, Gábor Lugosi, Gergely Neu. Random-Walk Perturbations for Online Combinatorial Optimization. IEEE Transactions on Information Theory, Institute of Electrical and Electronics Engineers, 2015, 61 (7), pp.4099 - 4106. ⟨10.1109/TIT.2015.2428253⟩. ⟨hal-01214987⟩

Share

Metrics

Record views

224

Files downloads

285