BOIL: Learning Environment Personalized Information
Provides a scalable method for multi-agent systems to extract environment information, improving long-horizon performance in complex environments.
BOIL extracts environment insights using PageRank and common information maximization, outperforming heuristic approaches in multi-agent tasks like coverage and patrolling over long time horizons.
Navigating complex environments poses challenges for multi-agent systems, requiring efficient extraction of insights from limited information. In this paper, we introduce the Blackbox Oracle Information Learning (BOIL) process, a scalable solution for extracting valuable insights from the environment structure. Leveraging the Pagerank algorithm and common information maximization, BOIL facilitates the extraction of information to guide long-term agent behavior applicable to problems such as coverage, patrolling, and stochastic reachability. Through experiments, we demonstrate the efficacy of BOIL in generating strategy distributions conducive to improved performance over extended time horizons, surpassing heuristic approaches in complex environments.