DC Field | Value | Language |
---|---|---|
dc.contributor.author | Kim, Jinhyeon | - |
dc.date.accessioned | 2023-06-22T19:31:10Z | - |
dc.date.available | 2023-06-22T19:31:10Z | - |
dc.date.issued | 2022 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=1000343&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/308176 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 김재철AI대학원, 2022.2,[iv, 19 p. :] | - |
dc.description.abstract | Text-based game is an instance of partially observable environment where the observation and action are in the form of natural language. Generalizing in text-based games serves as a useful stepping-stone towards reinforcement learning (RL) agent with generic linguistic ability. Prior works on generalization in RL often applied data augmentation techniques, but none of them focused on text-based games. We propose a novel data augmentation technique for text-based games, Transition-Matching Permutation, where we identify phrase permutations that match as many transitions in the trajectory data. Applying this technique resulted in the state-of-the-art performance in a procedurally generated TextWorld's Cooking Game benchmark. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | 텍스트 게임▼a자연어 처리▼a강화 학습▼a데이터 증강 | - |
dc.subject | Text-based game▼aNatural language processing▼aReinforcement learning▼aData augmentation | - |
dc.title | Data augmentation for learning to play in text-based games | - |
dc.title.alternative | 텍스트 게임에서의 일반화를 위한 데이터 증강 연구 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :김재철AI대학원, | - |
dc.contributor.alternativeauthor | 김진현 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.