Presentation 2016-11-17
Extraction of low dimensional attractors embedded in a recurrent neural network by using Dynamic Mode Decomposition
Shin Murata, Masato Okada,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Dynamic Mode Decomposition (DMD) decomposes high-dimensional dynamical data into a few dynamic modes and has been developed in the fluid-dynamics field. The decomposed modes consist of spatial bases and eigenvalues. We apply DMD to the simulation data of a recurrent neural network, and success to extract low-dimensional attractor embedded in the model.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Recurrent Neural Network / Dynamic Mode Decomposition
Paper # IBISML2016-85
Date of Issue 2016-11-09 (IBISML)

Conference Information
Committee IBISML
Conference Date 2016/11/16(3days)
Place (in Japanese) (See Japanese page)
Place (in English) Kyoto Univ.
Topics (in Japanese) (See Japanese page)
Topics (in English) Information-Based Induction Science Workshop (IBIS2016)
Chair Kenji Fukumizu(ISM)
Vice Chair Masashi Sugiyama(Univ. of Tokyo) / Hisashi Kashima(Kyoto Univ.)
Secretary Masashi Sugiyama(Univ. of Tokyo) / Hisashi Kashima(Nagoya Inst. of Tech.)
Assistant Toshihiro Kamishima(AIST) / Tomoharu Iwata(NTT)

Paper Information
Registration To Technical Committee on Infomation-Based Induction Sciences and Machine Learning
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Extraction of low dimensional attractors embedded in a recurrent neural network by using Dynamic Mode Decomposition
Sub Title (in English)
Keyword(1) Recurrent Neural Network
Keyword(2) Dynamic Mode Decomposition
1st Author's Name Shin Murata
1st Author's Affiliation The University of Tokyo(Univ. Tokyo)
2nd Author's Name Masato Okada
2nd Author's Affiliation The University of Tokyo/RIKEN(Univ. Tokyo/RIKEN)
Date 2016-11-17
Paper # IBISML2016-85
Volume (vol) vol.116
Number (no) IBISML-300
Page pp.pp.279-285(IBISML),
#Pages 7
Date of Issue 2016-11-09 (IBISML)