Combining Planning and Reinforcement Learning for Solving Relational Multiagent Domains

TL;DR

This paper introduces a method that combines relational planning with reinforcement learning to improve sample efficiency and generalization in complex multiagent environments, especially those with relational structures.

Contribution

It presents a novel integration of relational planners as centralized controllers with RL, addressing sample inefficiency and transferability in multiagent domains.

Findings

Enhanced sample efficiency in multiagent tasks

Improved transfer and generalization capabilities

Effective handling of relational domain complexities

Abstract

Multiagent Reinforcement Learning (MARL) poses significant challenges due to the exponential growth of state and action spaces and the non-stationary nature of multiagent environments. This results in notable sample inefficiency and hinders generalization across diverse tasks. The complexity is further pronounced in relational settings, where domain knowledge is crucial but often underutilized by existing MARL algorithms. To overcome these hurdles, we propose integrating relational planners as centralized controllers with efficient state abstractions and reinforcement learning. This approach proves to be sample-efficient and facilitates effective task transfer and generalization.