DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | 권인소 | - |
dc.contributor.author | Lee, Kyunghyun | - |
dc.contributor.author | 이경현 | - |
dc.date.accessioned | 2024-07-26T19:30:26Z | - |
dc.date.available | 2024-07-26T19:30:26Z | - |
dc.date.issued | 2023 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=1046592&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/320821 | - |
dc.description | 학위논문(박사) - 한국과학기술원 : 로봇공학학제전공, 2023.8,[vi, 55 p. :] | - |
dc.description.abstract | We introduce scalable Evolutionary Reinforcement Learning (ERL) Algorithms that combine Evolutionary Algorithm (EA) with Reinforcement Learning (RL) for better sample efficiency. It is widely known that the EA algorithms are less sample efficient than RL algorithms, but they are more stable. Combining two algorithms leads to more sample-efficient performance and stability. In this thesis, we introduce new asynchronous actor and critic update rules for scaling ERL algorithms and apply them to real-world applications where the sample efficiency is more crucial than simulated environments. The selected real-world environment is a camera parameter control task that is difficult to build a simulator. It is shown that the proposed ERL algorithm can achieve higher performance with less samples than conventional RL algorithms in both simulated and real-world environments. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | 진화론적 강화학습▼a강화학습▼a진화 전략▼a카메라 컨트롤 | - |
dc.subject | Evolutionary reinforcement learning▼aAsynchronous algorithm▼aEvolution strategy▼aCamera control | - |
dc.title | Practical evolutionary reinforcement learning with enhanced sample efficiency | - |
dc.title.alternative | 향상된 샘플 효율성을 통한 실제적인 진화론적 강화학습 | - |
dc.type | Thesis(Ph.D) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :로봇공학학제전공, | - |
dc.contributor.alternativeauthor | Kweon, In So | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.