자료유형 | 학위논문 |
---|---|
서명/저자사항 | A Picture of the Energy Landscape of Deep Neural Networks. |
개인저자 | Chaudhari, Pratik Anil. |
단체저자명 | University of California, Los Angeles. Computer Science 0201. |
발행사항 | [S.l.]: University of California, Los Angeles., 2018. |
발행사항 | Ann Arbor: ProQuest Dissertations & Theses, 2018. |
형태사항 | 175 p. |
기본자료 저록 | Dissertation Abstracts International 80-01B(E). Dissertation Abstract International |
ISBN | 9780438291706 |
학위논문주기 | Thesis (Ph.D.)--University of California, Los Angeles, 2018. |
일반주기 |
Source: Dissertation Abstracts International, Volume: 80-01(E), Section: B.
Adviser: Stefano Soatto. |
요약 | This thesis characterizes the training process of deep neural networks. We are driven by two apparent paradoxes. First, optimizing a non-convex function such as the loss function of a deep network should be extremely hard, yet rudimentary algori |
요약 | We build upon tools from two main areas to make progress on these questions: statistical physics and a continuous-time point-of-view of optimization. The former has been popular in the study of machine learning in the past and has been rejuvenat |
요약 | The confluence of these ideas leads to fundamental theoretical insights that explain observed phenomena in deep learning as well as the development of state-of-the-art algorithms for training deep networks. |
일반주제명 | Artificial intelligence. Applied mathematics. Statistical physics. |
언어 | 영어 |
바로가기 |
: 이 자료의 원문은 한국교육학술정보원에서 제공합니다. |