Thompson Sampling on Asymmetric $\alpha$-Stable Bandits

TL;DR

This paper explores the application of Thompson Sampling to multi-armed bandits with rewards following unknown asymmetric alpha-stable distributions, focusing on modeling financial and wireless data.

Contribution

It introduces the use of Thompson Sampling for bandits with asymmetric alpha-stable reward distributions, addressing a novel class of problems in reinforcement learning.

Findings

Demonstrates the effectiveness of Thompson Sampling in asymmetric alpha-stable bandit models

Provides insights into modeling financial and wireless data with alpha-stable distributions

Suggests potential applications in finance and wireless communication

Abstract

In algorithm optimization in reinforcement learning, how to deal with the exploration-exploitation dilemma is particularly important. Multi-armed bandit problem can optimize the proposed solutions by changing the reward distribution to realize the dynamic balance between exploration and exploitation. Thompson Sampling is a common method for solving multi-armed bandit problem and has been used to explore data that conform to various laws. In this paper, we consider the Thompson Sampling approach for multi-armed bandit problem, in which rewards conform to unknown asymmetric $\alpha$-stable distributions and explore their applications in modelling financial and wireless data.