Stochastic Constraint Programming as Reinforcement Learning

TL;DR

This paper introduces a hybrid approach combining Reinforcement Learning and Constraint Programming to effectively solve large stochastic constraint problems, leveraging the strengths of both methods.

Contribution

It presents a novel hybrid framework that integrates RL scalability with CP's modeling and filtering capabilities for stochastic constraint problems.

Findings

Prototype implementation demonstrates effectiveness on SCP problems.

Hybrid approach improves scalability over traditional SCP methods.

Shows potential for solving larger, more complex stochastic problems.

Abstract

Stochastic Constraint Programming (SCP) is an extension of Constraint Programming (CP) used for modelling and solving problems involving constraints and uncertainty. SCP inherits excellent modelling abilities and filtering algorithms from CP, but so far it has not been applied to large problems. Reinforcement Learning (RL) extends Dynamic Programming to large stochastic problems, but is problem-specific and has no generic solvers. We propose a hybrid combining the scalability of RL with the modelling and constraint filtering methods of CP. We implement a prototype in a CP system and demonstrate its usefulness on SCP problems.