2020年
Topological Measurement of Deep Neural Networks Using Persistent Homology.
International Symposium on Artificial Intelligence and Mathematics(ISAIM)
- ,
- 巻
- 90
- 号
- 1
- 開始ページ
- 75
- 終了ページ
- 92
- 記述言語
- 英語
- 掲載種別
- 研究論文(国際会議プロシーディングス)
- DOI
- 10.1007/s10472-021-09761-3
- 出版者・発行元
- SPRINGER
The inner representation of deep neural networks (DNNs) is indecipherable, which makes it difficult to tune DNN models, control their training process, and interpret their outputs. In this paper, we propose a novel approach to investigate the inner representation of DNNs through topological data analysis (TDA). Persistent homology (PH), one of the outstanding methods in TDA, was employed for investigating the complexities of trained DNNs. We constructed clique complexes on trained DNNs and calculated the one-dimensional PH of DNNs. The PH reveals the combinational effects of multiple neurons in DNNs at different resolutions, which is difficult to be captured without using PH. Evaluations were conducted using fully connected networks (FCNs) and networks combining FCNs and convolutional neural networks (CNNs) trained on the MNIST and CIFAR-10 data sets. Evaluation results demonstrate that the PH of DNNs reflects both the excess of neurons and problem difficulty, making PH one of the prominent methods for investigating the inner representation of DNNs.
- リンク情報
-
- DOI
- https://doi.org/10.1007/s10472-021-09761-3
- DBLP
- https://dblp.uni-trier.de/rec/conf/isaim/WatanabeY20
- Web of Science
- https://gateway.webofknowledge.com/gateway/Gateway.cgi?GWVersion=2&SrcAuth=JSTA_CEL&SrcApp=J_Gate_JST&DestLinkType=FullRecord&KeyUT=WOS:000669303900002&DestApp=WOS_CPL
- URL
- http://isaim2020.cs.ou.edu/papers/ISAIM2020_Watanabe_Yamana.pdf
- URL
- https://dblp.uni-trier.de/conf/isaim/2020
- URL
- https://dblp.uni-trier.de/db/conf/isaim/isaim2020.html#WatanabeY20
- Scopus
- https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85109252550&origin=inward 本文へのリンクあり
- Scopus Citedby
- https://www.scopus.com/inward/citedby.uri?partnerID=HzOxMe3b&scp=85109252550&origin=inward
- ID情報
-
- DOI : 10.1007/s10472-021-09761-3
- ISSN : 1012-2443
- eISSN : 1573-7470
- DBLP ID : conf/isaim/WatanabeY20
- SCOPUS ID : 85109252550
- Web of Science ID : WOS:000669303900002