arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.19090cs.LG
收藏

资源约束下的最优臂识别

Best Arm Identification with Resource Constraints

National University of Singapore(新加坡国立大学)

Zitian Li, Wang Chi Cheung

AI总结:

针对不同备选方案实验成本异质性问题,研究资源约束下的最优臂识别问题,提出带资源配给的连续减半算法SH-RR,实现接近最优的非渐近收敛速率,并发现确定性与随机资源消耗下收敛速率存在差异。

摘要:

受不同备选方案实验成本异质性的启发,我们研究了资源约束下的最优臂识别(Best Arm Identification with Resource Constraints,BAIwRC)问题。智能体的目标是在资源约束下识别最优臂,每次拉动臂都会消耗资源。我们做出了两项新颖贡献。我们设计并分析了带资源配给的连续减半算法(Successive Halving with Resource Rationing,SH-RR)。SH-RR在连续识别最优臂的概率方面,实现了接近最优的非渐近收敛速率。有趣的是,我们发现确定性资源消耗和随机资源消耗两种情形下的收敛速率存在差异。

英文摘要:

Motivated by the cost heterogeneity in experimentation across different alternatives, we study the Best Arm Identification with Resource Constraints (BAIwRC) problem. The agent aims to identify the best arm under resource constraints, where resources are consumed for each arm pull. We make two novel contributions. We design and analyze the Successive Halving with Resource Rationing algorithm (SH-RR). The SH-RR achieves a near-optimal non-asymptotic rate of convergence in terms of the probability of successively identifying an optimal arm. Interestingly, we identify a difference in convergence rates between the cases of deterministic and stochastic resource consumption.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑