Presentation 2006-06-16
Extraction of independent features in the hidden layer of Neural Networks
Kentaro MATSUKURA, Noboru MURATA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Neural Networks (NN) are simplified models of neural processing in a human brain. Its applications varies in wide areas of machine learning. Each output of the, unit consisting the hidden layer in a NN, is often similar when learning the network. On the other hand, nonlinear sparse coding can be performed by expressing the relation between the input and the output in the hidden layer. We can expect to extract information of the I/O from each unit of the hidden layer by this performance. We propose a method to extract independent features from the hidden layer in NNs and investigated it by a simple experiment.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Neural Network / Sparse Codeing
Paper # NC2006-33
Date of Issue

Conference Information
Committee NC
Conference Date 2006/6/9(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 JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Extraction of independent features in the hidden layer of Neural Networks
Sub Title (in English)
Keyword(1) Neural Network
Keyword(2) Sparse Codeing
1st Author's Name Kentaro MATSUKURA
1st Author's Affiliation Waseda University()
2nd Author's Name Noboru MURATA
2nd Author's Affiliation Waseda University
Date 2006-06-16
Paper # NC2006-33
Volume (vol) vol.106
Number (no) 102
Page pp.pp.-
#Pages 5
Date of Issue