arXiv research

Unified framework controls false discovery rate in bandit multiple testing.

PEAK tests means of multiple data streams with sequential betting.

We propose an adaptive sampling approach for multiple testing which aims to maximize statistical power while ensuring anytime false discovery control. We consider $n$ distributions whose means are partitioned by whether they are below or equal to a baseline (nulls), versus above the baseline (actual positives). In addi…

Simple greedy algorithms can excel in multi-objective bandits with multiple good arms.

Improves A/B testing power using a two-armed bandit framework.

New algorithm for efficiently identifying the best arm in stochastic bandits.

New algorithms improve exploration in unbounded reward settings.

Unified approach for non-stationary and clustered bandits.

Comparison Lift uses bandit algorithms to optimize online ad testing.

Study tests feasibility of linear programs with bandit feedback.

We consider a scenario where an agent has multiple available strategies to explore an unknown environment. For each new interaction with the environment, the agent must select which exploration strategy to use. We provide a new strategy-agnostic method that treat the situation as a Multi-Armed Bandits problem where the…

A UCB algorithm reduces regret in cooperative multi-agent graph bandits.

New methods improve LLM preference optimization by intelligently weighting multiple reference models.

Proposes a new TS algorithm for non-stationary bandits using KS tests.

Combines multiple bandit algorithms to create a nearly optimal single algorithm.

New methods for efficient exploration under unknown linear constraints in bandits.

Proposes DEXP3.M for unknown delay in multi-arm bandit with multiple play.

New algorithms improve model selection in linear bandits with optimal regret.

We study the problem of identifying the top $m$ arms in a multi-armed bandit game. Our proposed solution relies on a new algorithm based on successive rejects of the seemingly bad arms, and successive accepts of the good ones. This algorithmic contribution allows to tackle other multiple identifications settings that w…

We investigate the adversarial bandit problem with multiple plays under semi-bandit feedback. We introduce a highly efficient algorithm that asymptotically achieves the performance of the best switching $m$-arm strategy with minimax optimal regret bounds. To construct our algorithm, we introduce a new expert advice alg…

A new test improves statistical inference in bandit algorithms without sacrificing adaptiveness.

A new bandit algorithm for web page item display.

This work optimizes identifying good arms in nonparametric multi-armed bandits.

Deep Bayesian Bandits improve personalized ads by balancing exploration and exploitation.

New testing method for robust actor-critic bandit algorithms.

Data that is gathered adaptively --- via bandit algorithms, for example --- exhibits bias. This is true both when gathering simple numeric valued data --- the empirical means kept track of by stochastic bandit algorithms are biased downwards --- and when gathering more complicated data --- running hypothesis tests on c…

Optimizes COVID-19 testing policy using a Multi-Armed Bandit approach.

A new federated bandit problem with multiple adversaries, solved with a near-optimal algorithm.

As the cornerstone of modern portfolio theory, Markowitz's mean-variance optimization is considered a major model adopted in portfolio management. However, due to the difficulty of estimating its parameters, it cannot be applied to all periods. In some cases, naive strategies such as Equally-weighted and Value-weighted…

Unified meta-algorithm improves average performance across similar tasks in adversarial bandits.

Unified framework for robust A/B testing under model misspecification.

COBRA addresses strategic behavior in online platforms by ensuring truthful reporting without monetary incentives.

We consider the problem of best arm identification in a variant of multi-armed bandits called linked bandits. In a single interaction with linked bandits, multiple arms are played sequentially until one of them receives a positive reward. Since each interaction provides feedback about more than one arm, the sample comp…

Algorithm aggregates rewards from multiple players to learn related tasks in online bandit learning.

Optimizing news headlines is important for publishers and media sites. A compelling headline will increase readership, user engagement and social shares. At Yahoo Front Page, headline testing is carried out using a test-rollout strategy: we first allocate equal proportion of the traffic to each headline variation for a…

We present FIESTA, a model selection approach that significantly reduces the computational resources required to reliably identify state-of-the-art performance from large collections of candidate models. Despite being known to produce unreliable comparisons, it is still common practice to compare model evaluations base…

Algorithm improves multi-armed bandit performance by transferring reward samples.

This paper tackles combinatorial pure exploration for dueling bandits, aiming to find the best candidate-position match.

Improves A/B testing by detecting minor treatment effects.

We study stochastic multi-armed bandits with many players. The players do not know the number of players, cannot communicate with each other and if multiple players select a common arm they collide and none of them receive any reward. We consider the static scenario, where the number of players remains fixed, and the d…

Contextual bandits are online learners that, given an input, select an arm and receive a reward for that arm. They use the reward as a learning signal and aim to maximize the total reward over the inputs. Contextual bandits are commonly used to solve recommendation or ranking problems. This paper considers a learning s…

The paper tackles lifelong learning in multi-armed bandits, aiming to minimize average regret over multiple tasks.

We describe a novel algorithm for noisy global optimisation and continuum-armed bandits, with good convergence properties over any continuous reward function having finitely many polynomial maxima. Over such functions, our algorithm achieves square-root regret in bandits, and inverse-square-root error in optimisation, …

Paper tackles stochastic $k$-submodular bandits with full feedback, achieving sublinear regret.

New algorithm eliminates arms to minimize regret in complex bandit problems.

Survey and compare PAC-Bayes bounds for bandit problems.

Bayesian method infers contextual bandit policies robustly.

We consider the stochastic linear (multi-armed) contextual bandit problem with the possibility of hidden simple multi-armed bandit structure in which the rewards are independent of the contextual information. Algorithms that are designed solely for one of the regimes are known to be sub-optimal for the alternate regime…