English
Related papers

Related papers: Bayesian Regret Minimization in Offline Bandits

200 papers

This paper presents a finite-time analysis of the KL-UCB algorithm, an online, horizon-free index policy for stochastic bandit problems. We prove two distinct results: first, for arbitrary bounded rewards, the KL-UCB algorithm satisfies a…

Statistics Theory · Mathematics 2013-08-30 Aurélien Garivier , Olivier Cappé

Consider the sequential optimization of an expensive to evaluate and possibly non-convex objective function $f$ from noisy feedback, that can be considered as a continuum-armed bandit problem. Upper bounds on the regret performance of…

Machine Learning · Statistics 2021-03-11 Sattar Vakili , Kia Khezeli , Victor Picheny

We study stochastic linear bandits with heavy-tailed rewards, where the rewards have a finite $(1+\epsilon)$-absolute central moment bounded by $\upsilon$ for some $\epsilon \in (0,1]$. We improve both upper and lower bounds on the minimax…

Machine Learning · Computer Science 2026-01-28 Artin Tajdini , Jonathan Scarlett , Kevin Jamieson

We consider online convex optimization with a zero-order oracle feedback. In particular, the decision maker does not know the explicit representation of the time-varying cost functions, or their gradients. At each time step, she observes…

Optimization and Control · Mathematics 2020-05-05 Tatiana Tatarenko , Maryam Kamgarpour

Recent studies have shown that reinforcement learning with KL-regularized objectives can enjoy faster rates of convergence or logarithmic regret, in contrast to the classical $\sqrt{T}$-type regret in the unregularized setting. However, the…

Machine Learning · Computer Science 2026-03-03 Kaixuan Ji , Qingyue Zhao , Heyang Zhao , Qiwei Di , Quanquan Gu

We study best-arm identification (BAI) in the fixed-budget setting. Adaptive allocations based on upper confidence bounds (UCBs), such as UCBE, are known to work well in BAI. However, it is well-known that its optimal regret is…

Machine Learning · Computer Science 2024-10-24 Rong J. B. Zhu , Yanqi Qiu

We tackle the problem of online reward maximisation over a large finite set of actions described by their contexts. We focus on the case when the number of actions is too big to sample all of them even once. However we assume that we have…

Machine Learning · Computer Science 2013-09-27 Michal Valko , Nathaniel Korda , Remi Munos , Ilias Flaounas , Nelo Cristianini

In online inverse linear optimization, a learner observes time-varying sets of feasible actions and an agent's optimal actions, selected by solving linear optimization over the feasible actions. The learner sequentially makes predictions of…

Machine Learning · Computer Science 2025-05-23 Shinsaku Sakaue , Taira Tsuchiya , Han Bao , Taihei Oki

Bandits with Knapsacks (BwK) is a general model for multi-armed bandits under supply/budget constraints. While worst-case regret bounds for BwK are well-understood, we present three results that go beyond the worst-case perspective. First,…

Machine Learning · Computer Science 2021-12-30 Karthik Abinav Sankararaman , Aleksandrs Slivkins

Bayesian Optimization is critically vulnerable to extreme outliers. Existing provably robust methods typically assume a bounded cumulative corruption budget, which makes them defenseless against even a single corruption of sufficient…

Machine Learning · Statistics 2026-02-17 Abdelhamid Ezzerg , Ilija Bogunovic , Jeremias Knoblauch

We provide a simple method to combine stochastic bandit algorithms. Our approach is based on a "meta-UCB" procedure that treats each of $N$ individual bandit algorithms as arms in a higher-level $N$-armed bandit problem that we solve with a…

Machine Learning · Computer Science 2020-12-25 Ashok Cutkosky , Abhimanyu Das , Manish Purohit

In this paper, we analyze the problem of online convex optimization in different settings, including different feedback types (full-information/semi-bandit/bandit/etc) in either stochastic or non-stochastic setting and different notions of…

Machine Learning · Computer Science 2026-02-23 Mohammad Pedramfar , Vaneet Aggarwal

Traditional online learning models are typically initialized from scratch. By contrast, contemporary real-world applications often have access to historical datasets that can potentially enhanced the online learning processes. We study how…

Machine Learning · Computer Science 2025-12-19 Wang Chi Cheung , Lixing Lyu

We consider the Multi-Armed Bandit (MAB) problem, where an agent sequentially chooses actions and observes rewards for the actions it took. While the majority of algorithms try to minimize the regret, i.e., the cumulative difference between…

Machine Learning · Computer Science 2021-09-14 Nadav Merlis , Shie Mannor

Originally motivated by default risk management applications, this paper investigates a novel problem, referred to as the profitable bandit problem here. At each step, an agent chooses a subset of the K possible actions. For each action…

Machine Learning · Statistics 2018-05-09 Mastane Achab , Stephan Clémençon , Aurélien Garivier

Many physical systems have underlying safety considerations that require that the strategy deployed ensures the satisfaction of a set of constraints. Further, often we have only partial information on the state of the system. We study the…

Machine Learning · Computer Science 2022-03-30 Jiabin Lin , Xian Yeow Lee , Talukder Jubery , Shana Moothedath , Soumik Sarkar , Baskar Ganapathysubramanian

Dueling bandits is a prominent framework for decision-making involving preferential feedback, a valuable feature that fits various applications involving human interaction, such as ranking, information retrieval, and recommendation systems.…

Machine Learning · Computer Science 2024-10-16 Qiwei Di , Tao Jin , Yue Wu , Heyang Zhao , Farzad Farnoud , Quanquan Gu

In bandit settings, optimizing long-term regret metrics requires exploration, which corresponds to sometimes taking myopically sub-optimal actions. When a long-lived principal merely recommends actions to be executed by a sequence of…

Computer Science and Game Theory · Computer Science 2026-02-25 Ramya Ramalingam , Osbert Bastani , Aaron Roth

The contextual combinatorial semi-bandit problem with linear payoff functions is a decision-making problem in which a learner chooses a set of arms with the feature vectors in each round under given constraints so as to maximize the sum of…

Designing efficient general-purpose contextual bandit algorithms that work with large -- or even continuous -- action spaces would facilitate application to important scenarios such as information retrieval, recommendation systems, and…

Machine Learning · Computer Science 2022-07-14 Yinglun Zhu , Paul Mineiro