TY - GEN
T1 - Improving Generalization in Reinforcement Learning Through Forked Agents
AU - Moulin, Olivier
AU - Francois-Lavet, Vincent
AU - Elbers, Paul
AU - Hoogendoorn, Mark
N1 - Funding Information: A preprint version (earlier iteration) of this paper, written by the same authors can be found on arXiv: Moulin et al. [23]). Publisher Copyright: © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2023
Y1 - 2023
N2 - An eco-system of agents, each having their own policy with limited generalizability, has proven to be a reliable approach to increase generalization across procedurally generated environments. In such an approach, new agents are regularly added to the eco-system when encountering a new environment that is outside of the scope of the eco-system. The speed of adaptation and general effectiveness of the eco-system approach highly depends on the initialization of new agents. In this paper we propose different initialization techniques, inspired from Deep Neural Network initialization and transfer learning, and study their impact.
AB - An eco-system of agents, each having their own policy with limited generalizability, has proven to be a reliable approach to increase generalization across procedurally generated environments. In such an approach, new agents are regularly added to the eco-system when encountering a new environment that is outside of the scope of the eco-system. The speed of adaptation and general effectiveness of the eco-system approach highly depends on the initialization of new agents. In this paper we propose different initialization techniques, inspired from Deep Neural Network initialization and transfer learning, and study their impact.
KW - Agents
KW - Generalization
KW - Reinforcement Learning
UR - http://www.scopus.com/inward/record.url?scp=85169051103&partnerID=8YFLogxK
U2 - https://doi.org/10.1007/978-3-031-36822-6_22
DO - https://doi.org/10.1007/978-3-031-36822-6_22
M3 - Conference contribution
SN - 9783031368219
VL - 13926 LNAI
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 249
EP - 260
BT - Advances and Trends in Artificial Intelligence. Theory and Applications - 36th International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2023, Proceedings, Part II
A2 - Fujita, Hamido
A2 - Wang, Yinglin
A2 - Xiao, Yanghua
A2 - Moonis, Ali
PB - Springer Science and Business Media Deutschland GmbH
T2 - 36th International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2023
Y2 - 19 July 2023 through 22 July 2023
ER -