Stochastic bandits robust to adversarial corruptions

TL;DR

This paper introduces a new stochastic bandit model resilient to adversarial corruptions, proposing an algorithm that maintains near-optimal performance in stochastic settings and degrades gracefully with adversarial interference.

Contribution

It presents a simple, adaptive bandit algorithm that is robust to adversarial corruptions and provides theoretical guarantees on its performance degradation.

Findings

Algorithm's performance degrades linearly with corruption level C

Retains near-optimal guarantees in purely stochastic settings

Lower bound shows linear degradation is necessary for optimal stochastic performance

Abstract

We introduce a new model of stochastic bandits with adversarial corruptions which aims to capture settings where most of the input follows a stochastic pattern but some fraction of it can be adversarially changed to trick the algorithm, e.g., click fraud, fake reviews and email spam. The goal of this model is to encourage the design of bandit algorithms that (i) work well in mixed adversarial and stochastic models, and (ii) whose performance deteriorates gracefully as we move from fully stochastic to fully adversarial models. In our model, the rewards for all arms are initially drawn from a distribution and are then altered by an adaptive adversary. We provide a simple algorithm whose performance gracefully degrades with the total corruption the adversary injected in the data, measured by the sum across rounds of the biggest alteration the adversary made in the data in that round; this total corruption is denoted by $C$. Our algorithm provides a guarantee that retains the optimal guarantee (up to a logarithmic term) if the input is stochastic and whose performance degrades linearly to the amount of corruption $C$, while crucially being agnostic to it. We also provide a lower bound showing that this linear degradation is necessary if the algorithm achieves optimal performance in the stochastic setting (the lower bound works even for a known amount of corruption, a special case in which our algorithm achieves optimal performance without the extra logarithm).