DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Lee, Sang Wan | - |
dc.contributor.advisor | 이상완 | - |
dc.contributor.author | Kim, Heejun | - |
dc.date.accessioned | 2023-06-23T19:30:48Z | - |
dc.date.available | 2023-06-23T19:30:48Z | - |
dc.date.issued | 2023 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=1032731&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/308729 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 바이오및뇌공학과, 2023.2,[iv, 47 p. :] | - |
dc.description.abstract | One of the limitations of reinforcement learning (RL) algorithms is poor task generalizability. On the other hand, humans have the propensity to generalize environmental representations. This study aims to design a human-like generalizable RL algorithm using successor representation (SR), a computational model forming the human predictive map. We propose a novel method to quantify the invariance of the SR and show that it achieves environmental transformation invariance. Second, we implement an SR-Transformer model for task transfer, which best uses the SR's invariance. The proposed model outperforms baseline models on a zero-shot navigation task. We also demonstrate our model's generalizability on an image-based spatial navigation task. Critically, our model can explain various biological phenomena in memory-related brain areas, including the entorhinal grid and hippocampal place cells. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | Reinforcement learning▼aTransfer learning▼aSuccessor representation▼aPredictive map▼aGeneralization | - |
dc.subject | 강화학습▼a전이 학습▼a승계 표상▼a예측 지도▼a일반화 | - |
dc.title | (A) study on hippocampal successor representation for transfer learning | - |
dc.title.alternative | 전이 학습을 위한 해마의 승계 표상 연구 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :바이오및뇌공학과, | - |
dc.contributor.alternativeauthor | 김희준 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.