Better Algorithms for Stochastic Bandits with Adversarial Corruptions
classification
💻 cs.LG
stat.ML
keywords
adversarialalgorithmbanditscorruptionproblemstochasticagnosticalgorithms
read the original abstract
We study the stochastic multi-armed bandits problem in the presence of adversarial corruption. We present a new algorithm for this problem whose regret is nearly optimal, substantially improving upon previous work. Our algorithm is agnostic to the level of adversarial contamination and can tolerate a significant amount of corruption with virtually no degradation in performance.
This paper has not been read by Pith yet.
Forward citations
Cited by 1 Pith paper
-
Do Not Trust The Auctioneer: Learning to Bid in Feedback-Manipulated Auctions
In first-price auctions with feedback-only shilling, an algorithm combining robust interval elimination and optimistic debiasing with racing achieves near-optimal regret rates of O(T^{2/3}) or O(sqrt(T)) and matches a...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.