arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1603.02752cs.LGstat.ML
收藏

最佳K盗贼

Best-of-K Bandits

University of California, Berkeley(加州大学伯克利分校)

Max Simchowitz, Kevin Jamieson, Benjamin Recht

摘要:

本文研究了Best-of-K Bandit博弈:每次局中人在所有N个选项中选择一个子集S,并观察到奖励max(X(i):i in S),其中X是从联合分布中抽取的随机向量。我们的目标是确定的子集,实现最高的预期回报与高概率使用尽可能少的查询。我们提出了基于一个特定结构的依赖于分布的下界,该结构迫使学习者考虑所有N-choose-K子集,并匹配通过将每个子集视为单独的手臂而获得的bandit设置中已知上界的朴素扩展。然而,我们提出的证据表明,穷举搜索可以避免某些,有利的分布,因为高阶相关性的影响可以由低阶统计量支配。最后,我们提出了一个算法和分析独立的武器,它减轻了令人惊讶的非平凡的信息闭塞,发生由于只观察到的最大值的子集。这可能会通知战略更一般的依赖措施,我们补充这些结果与独立臂下限。

英文摘要:

This paper studies the Best-of-K Bandit game: At each time the player chooses a subset S among all N-choose-K possible options and observes reward max(X(i) : i in S) where X is a random vector drawn from a joint distribution. The objective is to identify the subset that achieves the highest expected reward with high probability using as few queries as possible. We present distribution-dependent lower bounds based on a particular construction which force a learner to consider all N-choose-K subsets, and match naive extensions of known upper bounds in the bandit setting obtained by treating each subset as a separate arm. Nevertheless, we present evidence that exhaustive search may be avoided for certain, favorable distributions because the influence of high-order order correlations may be dominated by lower order statistics. Finally, we present an algorithm and analysis for independent arms, which mitigates the surprising non-trivial information occlusion that occurs due to only observing the max in the subset. This may inform strategies for more general dependent measures, and we complement these result with independent-arm lower bounds.

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。

↑