Presentation 1997/12/18
Probabilistic Descent Method Applied to Similarity and Distance Measure of Quadratic Form for Pattern Recognition
Yoshiaki Kurosawa,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Probabilistic Descent (PD) was proposed by Amari and has been known as a method of the competitive learning in the statistical pattern recognition methods. This was expanded to GPD by Katagiri for flexible length vector data where the idea proposed by Amari was applied to fixed length vector data. The other competitive learning methods, LVQ, LSM, and ALSM were proposed by Kohonen and have been well known as effective methods for pattern recognition. Theoretical relationship between these methods are discussed in this report and the new approach for a learning method in this field is proposed.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Pattern Recognition / Character Recognition / PD / GPD / Probabilistic Descent
Paper # PRMU97-181
Date of Issue

Conference Information
Committee PRMU
Conference Date 1997/12/18(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 Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Probabilistic Descent Method Applied to Similarity and Distance Measure of Quadratic Form for Pattern Recognition
Sub Title (in English)
Keyword(1) Pattern Recognition
Keyword(2) Character Recognition
Keyword(3) PD
Keyword(4) GPD
Keyword(5) Probabilistic Descent
1st Author's Name Yoshiaki Kurosawa
1st Author's Affiliation Toshiba Research and Development Center()
Date 1997/12/18
Paper # PRMU97-181
Volume (vol) vol.97
Number (no) 458
Page pp.pp.-
#Pages 8
Date of Issue