DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Kim, Daeyoung | - |
dc.contributor.advisor | 김대영 | - |
dc.contributor.author | Nguyen, Van Giang | - |
dc.date.accessioned | 2021-05-13T19:38:27Z | - |
dc.date.available | 2021-05-13T19:38:27Z | - |
dc.date.issued | 2020 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=925167&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/285006 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 전산학부, 2020.8,[iii, 30 p. :] | - |
dc.description.abstract | Explaining the behaviors of deep neural networks, usually considered as black boxes, is critical especially when they are now being adopted over diverse aspects of human life. Taking the advantages of interpretable machine learning (interpretable ML), this work proposes a novel tool called Catastrophic Forgetting Dissector (or CFD) to explain catastrophic forgetting in continual learning settings. We also introduce a new method called Critical Freezing based on the observations of our tool. Experiments on ResNet articulate how catastrophic forgetting happens, particularly showing which components of this famous network are forgetting. Our new continual learning algorithm defeats various recent techniques by a significant margin, proving the capability of the investigation. Critical freezing not only attacks catastrophic forgetting but also exposes explainability. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | Image captioning▼aInterpretable ML▼aContinual learning▼aCatastrophic forgetting▼aDeep Visualization | - |
dc.subject | 이미지 캡션▼a해석 가능한 ML▼a지속적인 학습▼a파괴적 망각▼a깊은 시각화 | - |
dc.title | Overcoming catastrophic forgetting by deep visualization | - |
dc.title.alternative | 깊은 시각화를 이용한 파괴적 망각 극복 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :전산학부, | - |
dc.contributor.alternativeauthor | 뉴엔 반 지앙 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.