2013年5月
A neural network structure decomposition based on pruning and its visualization method
Journal of Advanced Computational Intelligence and Intelligent Informatics
- ,
- ,
- ,
- 巻
- 17
- 号
- 3
- 開始ページ
- 443
- 終了ページ
- 449
- 記述言語
- 英語
- 掲載種別
- 研究論文(学術雑誌)
- DOI
- 10.20965/jaciii.2013.p0443
- 出版者・発行元
- Fuji Technology Press
To decompose neural network structures for composite tasks, a pruning method and its visualization method are proposed. Visualization by placing the neurons on a 2D plane clarifies the structure related to each composited task. Experiments on a composite task using two tasks from a UCI dataset show that the neural network of the composite task contains more than 80% of neurons. The proposed methods target the transfer learning of robot motion, and results of an adaptation experiments are also referred. Copyright © 2013 Fuji Technology Press Co,. Ltd.
- リンク情報
- ID情報
-
- DOI : 10.20965/jaciii.2013.p0443
- ISSN : 1883-8014
- ISSN : 1343-0130
- eISSN : 1883-8014
- SCOPUS ID : 84879396910