Screening for an Infectious Disease as a Problem in Stochastic Control

TL;DR

This paper models infectious disease screening as a stochastic control problem, demonstrating that while optimal policies are complex, Thompson sampling offers provably optimal Bayesian regret guarantees, especially useful for poorly understood diseases like COVID-19.

Contribution

It introduces a stochastic-control framework for disease screening and proves the optimality of Thompson sampling in this context, addressing challenges in diseases with uncertain dynamics.

Findings

Thompson sampling achieves provably optimal Bayesian regret.

Optimal screening policies are provably difficult to compute.

Applicable to diseases with poorly understood dynamics like COVID-19.

Abstract

There has been much recent interest in screening populations for an infectious disease. Here, we present a stochastic-control model, wherein the optimum screening policy is provably difficult to find, but wherein Thompson sampling has provably optimal performance guarantees in the form of Bayesian regret. Thompson sampling seems applicable especially to diseases, for which we do not understand the dynamics well, such as to the super-spreading COVID-19.