Discrete-time Approximation of Stochastic Optimal Control with Partial Observation

TL;DR

This paper develops a convergent discrete-time approximation method for stochastic optimal control problems with partial observation, enabling practical numerical algorithms and potential integration with machine learning techniques.

Contribution

It introduces a new discrete-time approximation framework with proven convergence for partially observed stochastic control problems, facilitating numerical solutions.

Findings

Established convergence of discrete-time control approximations

Provided a practical numerical algorithm with convergence guarantees

Demonstrated the approach with linear quadratic numerical experiments

Abstract

We consider a class of stochastic optimal control problems with partial observation, and study their approximation by discrete-time control problems. We establish a convergence result by using weak convergence technique of Kushner and Dupuis [Numerical Methods for Stochastic Control Problems in Continuous Time (2001), Springer-Verlag, New York], together with the notion of relaxed control rule introduced by El Karoui, Huu Nguyen and Jeanblanc-Picqu\'e [SIAM J. Control Optim., 26 (1988) 1025-1061]. In particular, with a well chosen discrete-time control system, we obtain a first implementable numerical algorithm (with convergence) for the partially observed control problem. Moreover, our discrete-time approximation result would open the door to study convergence of more general numerical approximation methods, such as machine learning based methods. Finally, we illustrate our convergence result by the numerical experiments on a partially observed control problem in a linear quadratic setting.