Stochastic Gradient Succeeds for Bandits

• Published in: arXiv.org
• Main Authors: Mei, Jincheng, Zhong, Zixin, Dai, Bo, Agarwal, Alekh, Szepesvari, Csaba, Schuurmans, Dale
• Format: Paper
• Language: English
• Published: Ithaca Cornell University Library, arXiv.org 27.02.2024
• Subjects: Algorithms; Convergence; Noise control; Optimization

We show that the \emph{stochastic gradient} bandit algorithm converges to a \emph{globally optimal} policy at an \(O(1/t)\) rate, even with a \emph{constant} step size. Remarkably, global convergence of the stochastic gradient bandit algorithm has not been previously established, even though it is a...