arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2107.06196cs.LG
收藏

第1001章学土匪不后悔

No Regrets for Learning the Prior in Bandits

Google(谷歌);Google Research(谷歌研究院);DeepMind(深度思维);University of Alberta(阿尔伯塔大学)

Soumya Basu, Branislav Kveton, Manzil Zaheer, Csaba Szepesvári

摘要:

我们提出了${\tt AdaTS}$,汤普森采样算法,适应顺序的强盗任务,它与互动。${\tt AdaTS}$的核心思想是通过保持任务参数的分布来适应未知的任务先验分布。在解决一个强盗任务时,这种不确定性被边缘化并得到适当的解释。${\tt AdaTS}$是一个完全贝叶斯算法,可以有效地实现在几类强盗问题。我们推导出其贝叶斯遗憾的上限,量化的损失,由于不知道的任务之前,并表明它是小的。我们的理论得到了实验的支持,其中${\tt AdaTS}$优于以前的算法,即使在具有挑战性的现实问题中也能很好地工作。

英文摘要:

We propose ${\tt AdaTS}$, a Thompson sampling algorithm that adapts sequentially to bandit tasks that it interacts with. The key idea in ${\tt AdaTS}$ is to adapt to an unknown task prior distribution by maintaining a distribution over its parameters. When solving a bandit task, that uncertainty is marginalized out and properly accounted for. ${\tt AdaTS}$ is a fully-Bayesian algorithm that can be implemented efficiently in several classes of bandit problems. We derive upper bounds on its Bayes regret that quantify the loss due to not knowing the task prior, and show that it is small. Our theory is supported by experiments, where ${\tt AdaTS}$ outperforms prior algorithms and works well even in challenging real-world problems.

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。

↑