Approximate Control for Continuous-Time POMDPs

TL;DR

This paper introduces an approximation framework for decision-making in continuous-time POMDPs with discrete states, enabling scalable control solutions by projecting high-dimensional filtering distributions onto parametric families.

Contribution

It presents a novel scalable control heuristic that combines parametric filtering with fully observable system control for continuous-time POMDPs.

Findings

Effective in queueing systems

Applicable to chemical reaction networks

Scales well with large state spaces

Abstract

This work proposes a decision-making framework for partially observable systems in continuous time with discrete state and action spaces. As optimal decision-making becomes intractable for large state spaces we employ approximation methods for the filtering and the control problem that scale well with an increasing number of states. Specifically, we approximate the high-dimensional filtering distribution by projecting it onto a parametric family of distributions, and integrate it into a control heuristic based on the fully observable system to obtain a scalable policy. We demonstrate the effectiveness of our approach on several partially observed systems, including queueing systems and chemical reaction networks.