Low-rank Tensor Bandits
Low-rank Tensor Bandits
复制标题
低阶张量强盗
DOI：
--
复制
发表时间：
2020
期刊：
影响因子：
--
通讯作者：
W. Sun
中科院分区：
文献类型：
--
作者：
Botao Hao;Jie Zhou;Zheng Wen;W. Sun
近年来，多维在线决策在许多实际应用中发挥着至关重要的作用，如在线推荐和数字营销。为了解决这个问题，我们引入随机低秩张量强盗，一类强盗的平均奖励可以表示为一个低秩张量。我们提出了两个学习算法，张量epoch-greedy和张量消除，并开发有限时间遗憾界。我们观察到，张量消除有一个最佳的依赖于时间范围，而张量epoch-greedy有一个更尖锐的依赖于张量尺寸。数值实验进一步支持了这些理论研究结果，并表明我们的算法优于各种忽略张量低秩结构的最先进的方法。
In recent years, multi-dimensional online decision making has been playing a crucial role in many practical applications such as online recommendation and digital marketing. To solve it, we introduce stochastic low-rank tensor bandits, a class of bandits whose mean rewards can be represented as a low-rank tensor. We propose two learning algorithms, tensor epoch-greedy and tensor elimination, and develop finite-time regret bounds for them. We observe that tensor elimination has an optimal dependency on the time horizon, while tensor epoch-greedy has a sharper dependency on tensor dimensions. Numerical experiments further back up these theoretical findings and show that our algorithms outperform various state-of-the-art approaches that ignore the tensor low-rank structure.
影响因子：
2.5
作者：
Ming Yuan;Cun-Hui Zhang
通讯作者：
Ming Yuan;Cun-Hui Zhang