Presentation 2009-07-13
Learning to imitate stochastic time series in a compositional way by chaos
Jun NAMIKAWA, Jun TANI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This study shows that a mixture of RNN experts model can acquire the ability to generate sequences that are combination of multiple primitive patterns by means of self-organizing chaos. By training of the model, each expert learns a primitive sequence pattern, and a gating network learns to imitate stochastic switching of the multiple primitives via a chaotic dynamics, utilizing a sensitive dependence on initial conditions. As a demonstration, we present a numerical simulation in which the model learns Markov chain switching among some Lissajous curves by a chaotic dynamics. Our analysis shows that a self-organized chaotic system can reconstruct the probability of primitive switching as observed in the training data.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) recurrent neural network / mixture of experts / maximum likelihood estimation / chaos
Paper # NLP2009-18,NC2009-11
Date of Issue

Conference Information
Committee NC
Conference Date 2009/7/6(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Learning to imitate stochastic time series in a compositional way by chaos
Sub Title (in English)
Keyword(1) recurrent neural network
Keyword(2) mixture of experts
Keyword(3) maximum likelihood estimation
Keyword(4) chaos
1st Author's Name Jun NAMIKAWA
1st Author's Affiliation Brain Science Institute, RIKEN()
2nd Author's Name Jun TANI
2nd Author's Affiliation Brain Science Institute, RIKEN
Date 2009-07-13
Paper # NLP2009-18,NC2009-11
Volume (vol) vol.109
Number (no) 125
Page pp.pp.-
#Pages 6
Date of Issue