paper-with-me

홈 › Papers

MaxGap Bandit: Adaptive Algorithms for Approximate Ranking

2019-06-03 · NeurIPS 2019 12 · Sumeet Katariya, Ardhendu Tripathy, Robert Nowak

This paper studies the problem of adaptively sampling from K distributions (arms) in order to identify the largest gap between any two adjacent means. We call this the MaxGap-bandit problem. This problem arises naturally in approximate ranking, noisy sorting, outlier detection, and top-arm identification in bandits. The key novelty of the MaxGap-bandit problem is that it aims to adaptively determine the natural partitioning of the distributions into a subset with larger means and a subset with smaller means, where the split is determined by the largest gap rather than a pre-specified rank or threshold. Estimating an arm's gap requires sampling its neighboring arms in addition to itself, and this dependence results in a novel hardness parameter that characterizes the sample complexity of the problem. We propose elimination and UCB-style algorithms and show that they are minimax optimal. Our experiments show that the UCB-style algorithms require 6-8x fewer samples than non-adaptive sampling to achieve the same error.

📄 PDF Abstract BibTeX arXiv:1906.00547

Code (1)

sumeetsk/maxgap_bandit 공식 구현

Tasks

Outlier Detection

Similar Papers 제목 키워드 기반

Eliciting Kemeny Rankings

2023-12-18 · Anne-Marie George, Christos Dimitrakakis

We formulate the problem of eliciting agents' preferences with the goal of finding a Kemeny ranking as a Dueling Bandits problem. Here the bandits' arms correspond to alternatives that need to be ranked and the feedback …

Adversarial Online Learning with Changing Action Sets: Efficient Algorithms with Approximate Regret Bounds

2020-03-07 · Ehsan Emamjomeh-Zadeh, Chen-Yu Wei, Haipeng Luo, David Kempe

We revisit the problem of online learning with sleeping experts/bandits: in each time step, only a subset of the actions are available for the algorithm to choose from (and learn about). The work of Kleinberg et al. (201…

PAC learning

An Asymptotically Optimal Batched Algorithm for the Dueling Bandit Problem

2022-09-25 · Arpit Agarwal, Rohan Ghuge, Viswanath Nagarajan

We study the $K$-armed dueling bandit problem, a variation of the traditional multi-armed bandit problem in which feedback is obtained in the form of pairwise comparisons. Previous learning algorithms have focused on the…

Recommendation Systems

Inference for Batched Bandits

2020-02-08 · NeurIPS 2020 12 · Kelly W. Zhang, Lucas Janson, Susan A. Murphy

As bandit algorithms are increasingly utilized in scientific studies and industrial applications, there is an associated increasing need for reliable inference methods based on the resulting adaptively-collected data. In…

Multi-Armed Bandits

Adapting multi-armed bandits policies to contextual bandits scenarios

2018-11-11 · David Cortes

This work explores adaptations of successful multi-armed bandits policies to the online contextual bandits scenario with binary rewards using binary classification algorithms such as logistic regression as black-box orac…

Binary ClassificationClassificationGeneral ClassificationMulti-Armed Bandits+2