arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2405.07331cs.LGcs.DSstat.ML
收藏

基于ReLU神经网络的随机多臂老虎机

Stochastic Bandits with ReLU Neural Networks

Arizona State University(亚利桑那州立大学);University of Pennsylvania(宾夕法尼亚大学)

Kan Xu, Hamsa Bastani, Surbhi Goel, Osbert Bastani

AI总结:

本研究针对带ReLU神经网络结构的随机多臂老虎机问题,提出OFU-ReLU及OFU-ReLU+算法,首次实现单层ReLU下$\tilde{O}(\sqrt{T})$的遗憾值上界,核心是利用ReLU分段线性结构将问题转化为变换特征空间的线性老虎机问题。

摘要:

我们研究具有ReLU神经网络结构的随机多臂老虎机问题。研究表明,针对单层ReLU神经网络的多臂老虎机,可实现$\tilde{O}(\sqrt{T})$的遗憾值上界;据我们所知,本研究是首个达成该保证的工作。在此特定设定下,我们提出OFU-ReLU算法以实现这一上界。该算法先进行随机探索直至进入线性区域,随后执行UCB型线性多臂老虎机算法以平衡探索与利用。我们的核心洞见是:一旦在探索阶段相对准确地学习到ReLU的参数,便可利用ReLU激活的分段线性结构,将问题转化为变换特征空间中的线性多臂老虎机问题。为消除对模型参数的依赖,我们基于批量策略设计了OFU-ReLU+算法,可提供相同的理论保证。

英文摘要:

We study the stochastic bandit problem with ReLU neural network structure. We show that a $\tilde{O}(\sqrt{T})$ regret guarantee is achievable by considering bandits with one-layer ReLU neural networks; to the best of our knowledge, our work is the first to achieve such a guarantee. In this specific setting, we propose an OFU-ReLU algorithm that can achieve this upper bound. The algorithm first explores randomly until it reaches a linear regime, and then implements a UCB-type linear bandit algorithm to balance exploration and exploitation. Our key insight is that we can exploit the piecewise linear structure of ReLU activations and convert the problem into a linear bandit in a transformed feature space, once we learn the parameters of ReLU relatively accurately during the exploration stage. To remove dependence on model parameters, we design an OFU-ReLU+ algorithm based on a batching strategy, which can provide the same theoretical guarantee.

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。

↑