2026-09-18T20:01:17Zhttps://dspace.mit.edu/server/oai/requestoai:dspace.mit.edu:1721.1/1391432022-01-15T04:00:23Zcom_1721.1_7582com_1721.1_7581col_1721.1_131023
Shah, Devavrat
Alumootil, Varkey
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
2022-01-14T14:52:30Z
2022-01-14T14:52:30Z
2021-06
2021-06-17T20:12:49.354Z
https://hdl.handle.net/1721.1/139143
Performance of state-of-the art offline and model-based reinforcement learning (RL) algorithms deteriorates significantly when subjected to severe data scarcity and the presence of heterogeneous agents. In this work, we propose a model-based offline RL method to approach this setting. Using all available data from the various agents, we construct personalized simulators for each individual agent, which are then used to train RL policies. We do so by modeling the transition dynamics of the agents as a low rank tensor decomposition of latent factors associated with agents, states, and actions. We perform experiments on various benchmark environments and demonstrate improvement over existing offline approaches in the scarce data regime.
M.Eng.
Massachusetts Institute of Technology
In Copyright - Educational Use Permitted
Copyright MIT
http://rightsstatements.org/page/InC-EDU/1.0/
Data-Efficient Offline Reinforcement Learning with Heterogeneous Agents
Thesis
application/pdf
Master
Master of Engineering in Electrical Engineering and Computer Science
Publication
unknown
unknown
<Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="27b92573-5e7a-4036-958a-ae6d721d04eb">
<Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type>
<Title>Data-Efficient Offline Reinforcement Learning with Heterogeneous Agents</Title>
<PublishedIn>
<Publication>
</Publication>
</PublishedIn>
<PublicationDate>2021-06</PublicationDate>
<Authors>
<Author>
<DisplayName>Alumootil, Varkey</DisplayName>
<Affiliation>
<OrgUnit>
</OrgUnit>
</Affiliation>
</Author>
</Authors>
<Editors>
</Editors>
<Publishers>
<Publisher>
<DisplayName>Massachusetts Institute of Technology</DisplayName>
<OrgUnit />
</Publisher>
</Publishers>
<License>http://rightsstatements.org/page/InC-EDU/1.0/</License>
<Abstract>Performance of state-of-the art offline and model-based reinforcement learning (RL) algorithms deteriorates significantly when subjected to severe data scarcity and the presence of heterogeneous agents. In this work, we propose a model-based offline RL method to approach this setting. Using all available data from the various agents, we construct personalized simulators for each individual agent, which are then used to train RL policies. We do so by modeling the transition dynamics of the agents as a low rank tensor decomposition of latent factors associated with agents, states, and actions. We perform experiments on various benchmark environments and demonstrate improvement over existing offline approaches in the scarce data regime.</Abstract>
<Access xmlns="http://purl.org/coar/access_right"
>
</Access>
</Publication>