Improving reinforcement learning with ensembles of different learners

Crafford, Gerhardus JRosman, B2023-01-272023-01-272022-11Crafford, G.J. & Rosman, B. 2022. Improving reinforcement learning with ensembles of different learners. http://hdl.handle.net/10204/12598 .https://doi.org/10.1051/matecconf/202237007008http://hdl.handle.net/10204/12598Different reinforcement learning (RL) methods exist to address the problem of combining multiple different learners to generate a superior learner. These existing methods usually assume that each learner uses the same algorithm and/or state representation. We propose an ensemble learner that combines a set of base learners and leverages the strengths of the different base learners online. We demonstrate the proposed ensemble learner’s ability to combine the strengths of multiple base learners and adapt to changes in base learner performance on various domains, including the Atari Breakout domain.FulltextenReinforcement learningEnsemble learningImproving reinforcement learning with ensembles of different learnersConference PresentationCrafford, G. J., & Rosman, B. (2022). Improving reinforcement learning with ensembles of different learners. http://hdl.handle.net/10204/12598Crafford, Gerhardus J, and B Rosman. "Improving reinforcement learning with ensembles of different learners." <i>23rd Annual International RAPDASA Conference joined by RobMech, PRASA and CoSAAMI, Somerset-West, Cape Town, 9-11 November 2022</i> (2022): http://hdl.handle.net/10204/12598Crafford GJ, Rosman B, Improving reinforcement learning with ensembles of different learners; 2022. http://hdl.handle.net/10204/12598 .TY - Conference Presentation AU - Crafford, Gerhardus J AU - Rosman, B AB - Different reinforcement learning (RL) methods exist to address the problem of combining multiple different learners to generate a superior learner. These existing methods usually assume that each learner uses the same algorithm and/or state representation. We propose an ensemble learner that combines a set of base learners and leverages the strengths of the different base learners online. We demonstrate the proposed ensemble learner’s ability to combine the strengths of multiple base learners and adapt to changes in base learner performance on various domains, including the Atari Breakout domain. DA - 2022-11 DB - ResearchSpace DP - CSIR J1 - 23rd Annual International RAPDASA Conference joined by RobMech, PRASA and CoSAAMI, Somerset-West, Cape Town, 9-11 November 2022 KW - Reinforcement learning KW - Ensemble learning LK - https://researchspace.csir.co.za PY - 2022 T1 - Improving reinforcement learning with ensembles of different learners TI - Improving reinforcement learning with ensembles of different learners UR - http://hdl.handle.net/10204/12598 ER -37073