DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Yang, Hongseok | - |
dc.contributor.advisor | 양홍석 | - |
dc.contributor.author | Park, Jeongmin | - |
dc.date.accessioned | 2022-04-27T19:32:03Z | - |
dc.date.available | 2022-04-27T19:32:03Z | - |
dc.date.issued | 2021 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=963362&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/296133 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 전산학부, 2021.8,[iii, 18 p. :] | - |
dc.description.abstract | Pruning is a technique that removes unnecessary weights from the neural network to obtain a memory efficient subnetwork with comparable accuracy to the original network. Lottery Ticket Hypothesis conjectures that most randomly-initialized, dense networks have subnetworks that match accuracy of the original networks when trained in isolation, and Iterative Magnitude Pruning (IMP) is the algorithm typically used to find such subnetworks. We conjecture that the Neural Tangnet Kernel (NTK) theory may be used to analyze Lottery Ticket Hypothesis and the IMP algorithm. We give an intuition about the conjecture, and provide an experimental evidence for our conjecture using the MNIST dataset. We also suggest a novel pruning scheme based on the conjecture. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | Overparameterization▼aPruning▼aWide neural network▼aLinear connectivity▼aLinear approximation▼aIterative magnitude pruning▼aLottery ticket hypothesis▼aNeural tangent kernel | - |
dc.subject | 과매개변수화▼a가지치기▼a넓은 신경망▼a선형 연결성▼a선형 근사 | - |
dc.title | On the connection between lottery ticket hypothesis and neural tangent Kernel | - |
dc.title.alternative | Lottery ticket hypothesis와 neural tangent kernel의 관계에 대한 연구 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :전산학부, | - |
dc.contributor.alternativeauthor | 박정민 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.