TY - JOUR
T1 - Learning to Operate in Open Worlds by Adapting Planning Models
AU - Piotrowski, Wiktor
AU - Stern, Roni
AU - Sher, Yoni
AU - Le, Jacob
AU - Klenk, Matthew
AU - deKleer, Johan
AU - Mohan, Shiwali
N1 - Funding Information:
The work presented in this paper was supported in part by the DARPA SAIL-ON program under award number HR001120C0040. The views, opinions and/or findings expressed are those of the authors' and should not be interpreted as representing the official views or policies of the Department of Defense or the U.S. Government.
Funding Information:
The work presented in this paper was supported in part by the DARPA SAIL-ON program under award number HR001120C0040. The views, opinions and/or findings expressed are those of the authors’ and should not be interpreted as representing the official views or policies of the Department of Defense or the U.S. Government.
Publisher Copyright:
© 2023 International Foundation for Autonomous Agents and Multiagent Systems (www.ifaamas.org). All rights reserved.
PY - 2023/1/1
Y1 - 2023/1/1
N2 - Planning agents are ill-equipped to act in novel situations in which their domain model no longer accurately represents the world. We introduce an approach for such agents operating in open worlds that detects the presence of novelties and effectively adapts their domain models and consequent action selection. It uses observations of action execution and measures their divergence from what is expected, according to the environment model, to infer existence of a novelty. Then, it revises the model through a heuristics-guided search over model changes. We report empirical evaluations on the CartPole problem, a standard Reinforcement Learning (RL) benchmark. The results show that our approach can deal with a class of novelties very quickly and in an interpretable fashion.
AB - Planning agents are ill-equipped to act in novel situations in which their domain model no longer accurately represents the world. We introduce an approach for such agents operating in open worlds that detects the presence of novelties and effectively adapts their domain models and consequent action selection. It uses observations of action execution and measures their divergence from what is expected, according to the environment model, to infer existence of a novelty. Then, it revises the model through a heuristics-guided search over model changes. We report empirical evaluations on the CartPole problem, a standard Reinforcement Learning (RL) benchmark. The results show that our approach can deal with a class of novelties very quickly and in an interpretable fashion.
KW - Adaptive Agents
KW - Model Repair
KW - Open World Learning
KW - Planning
UR - http://www.scopus.com/inward/record.url?scp=85171291182&partnerID=8YFLogxK
M3 - Conference article
AN - SCOPUS:85171291182
SN - 1548-8403
VL - 2023-May
SP - 2610
EP - 2612
JO - Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems, AAMAS
JF - Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems, AAMAS
T2 - 22nd International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2023
Y2 - 29 May 2023 through 2 June 2023
ER -