Evolutionary Multi-Armed Bandits with Genetic Thompson Sampling

TL;DR

This paper introduces Genetic Thompson Sampling, a novel bandit algorithm that integrates genetic algorithms to enhance decision-making in nonstationary environments, demonstrated through simulations and epidemic control applications.

Contribution

It presents the first integration of evolutionary principles into multi-armed bandit algorithms, specifically through Genetic Thompson Sampling, improving performance in dynamic settings.

Findings

Outperforms baseline algorithms in nonstationary environments

Effective in epidemic control simulation

Provides an interactive visualization tool for learning process

Abstract

As two popular schools of machine learning, online learning and evolutionary computations have become two important driving forces behind real-world decision making engines for applications in biomedicine, economics, and engineering fields. Although there are prior work that utilizes bandits to improve evolutionary algorithms' optimization process, it remains a field of blank on how evolutionary approach can help improve the sequential decision making tasks of online learning agents such as the multi-armed bandits. In this work, we propose the Genetic Thompson Sampling, a bandit algorithm that keeps a population of agents and update them with genetic principles such as elite selection, crossover and mutations. Empirical results in multi-armed bandit simulation environments and a practical epidemic control problem suggest that by incorporating the genetic algorithm into the bandit algorithm, our method significantly outperforms the baselines in nonstationary settings. Lastly, we introduce EvoBandit, a web-based interactive visualization to guide the readers through the entire learning process and perform lightweight evaluations on the fly. We hope to engage researchers into this growing field of research with this investigation.