arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2106.01660stat.MLcs.LGmath.STstat.MEstat.TH
收藏

Bandit相位恢复

Bandit Phase Retrieval

DeepMind(深度思维)

Tor Lattimore, Botao Hao

摘要:

我们研究了一个强盗版本的相位检索,学习者选择行动$(A_t)_{t=1}^n$在$d$维单位球和预期的奖励是$\langle A_t, θ_\star\rangle^2$,其中$θ_\star \in \mathbb R^d$是一个未知的参数向量。我们证明了在这个问题中的极大极小累积后悔是$\smash{\tilde Θ(d \sqrt{n})}$,它提高了最好的已知的界限的一个因素$\smash{\sqrt{d}}$。我们还表明,极小极大简单的遗憾是$\smash{\tilde Θ(d / \sqrt{n})}$,这是唯一可实现的自适应算法。我们的分析表明,一个明显令人信服的启发式猜测下限可能是误导性的,信息导向抽样的信息比的统一界限是不够的最佳遗憾。

英文摘要:

We study a bandit version of phase retrieval where the learner chooses actions $(A_t)_{t=1}^n$ in the $d$-dimensional unit ball and the expected reward is $\langle A_t, θ_\star\rangle^2$ where $θ_\star \in \mathbb R^d$ is an unknown parameter vector. We prove that the minimax cumulative regret in this problem is $\smash{\tilde Θ(d \sqrt{n})}$, which improves on the best known bounds by a factor of $\smash{\sqrt{d}}$. We also show that the minimax simple regret is $\smash{\tilde Θ(d / \sqrt{n})}$ and that this is only achievable by an adaptive algorithm. Our analysis shows that an apparently convincing heuristic for guessing lower bounds can be misleading and that uniform bounds on the information ratio for information-directed sampling are not sufficient for optimal regret.

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。

↑