TY - GEN
T1 - PAnDR
T2 - 31st International Joint Conference on Artificial Intelligence, IJCAI 2022
AU - Sang, Tong
AU - Tang, Hongyao
AU - Ma, Yi
AU - Hao, Jianye
AU - Zheng, Yan
AU - Meng, Zhaopeng
AU - Li, Boyan
AU - Wang, Zhen
N1 - Publisher Copyright:
© 2022 International Joint Conferences on Artificial Intelligence. All rights reserved.
PY - 2022
Y1 - 2022
N2 - Deep Reinforcement Learning (DRL) has been a promising solution to many complex decision-making problems. Nevertheless, the notorious weakness in generalization among environments prevent widespread application of DRL agents in real-world scenarios. Although advances have been made recently, most prior works assume sufficient online interaction on training environments, which can be costly in practical cases. To this end, we focus on an offline-training-online-adaptation setting, in which the agent first learns from offline experiences collected in environments with different dynamics and then performs online policy adaptation in environments with new dynamics. In this paper, we propose Policy Adaptation with Decoupled Representations (PAnDR) for fast policy adaptation. In offline training phase, the environment representation and policy representation are learned through contrastive learning and policy recovery, respectively. The representations are further refined by mutual information optimization to make them more decoupled and complete. With learned representations, a Policy-Dynamics Value Function (PDVF) [Raileanu et al., 2020] network is trained to approximate the values for different combinations of policies and environments from offline experiences. In online adaptation phase, with the environment context inferred from few experiences collected in new environments, the policy is optimized by gradient ascent with respect to the PDVF. Our experiments show that PAnDR outperforms existing algorithms in several representative policy adaptation problems.
AB - Deep Reinforcement Learning (DRL) has been a promising solution to many complex decision-making problems. Nevertheless, the notorious weakness in generalization among environments prevent widespread application of DRL agents in real-world scenarios. Although advances have been made recently, most prior works assume sufficient online interaction on training environments, which can be costly in practical cases. To this end, we focus on an offline-training-online-adaptation setting, in which the agent first learns from offline experiences collected in environments with different dynamics and then performs online policy adaptation in environments with new dynamics. In this paper, we propose Policy Adaptation with Decoupled Representations (PAnDR) for fast policy adaptation. In offline training phase, the environment representation and policy representation are learned through contrastive learning and policy recovery, respectively. The representations are further refined by mutual information optimization to make them more decoupled and complete. With learned representations, a Policy-Dynamics Value Function (PDVF) [Raileanu et al., 2020] network is trained to approximate the values for different combinations of policies and environments from offline experiences. In online adaptation phase, with the environment context inferred from few experiences collected in new environments, the policy is optimized by gradient ascent with respect to the PDVF. Our experiments show that PAnDR outperforms existing algorithms in several representative policy adaptation problems.
UR - http://www.scopus.com/inward/record.url?scp=85137879187&partnerID=8YFLogxK
U2 - 10.24963/ijcai.2022/474
DO - 10.24963/ijcai.2022/474
M3 - 会议稿件
AN - SCOPUS:85137879187
T3 - IJCAI International Joint Conference on Artificial Intelligence
SP - 3416
EP - 3422
BT - Proceedings of the 31st International Joint Conference on Artificial Intelligence, IJCAI 2022
A2 - De Raedt, Luc
A2 - De Raedt, Luc
PB - International Joint Conferences on Artificial Intelligence
Y2 - 23 July 2022 through 29 July 2022
ER -