Presentation 2019-10-12
An Optimization for Classification by Self-Organizing Maps Based on Attribute Information
Tetsuya Sato, Kazuma Tsuchida, Yukari Yamauti,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Self-Organizing Map (SOM) is a simple algorithm that has excellent clustering capabilities and adapts continuous changes. However, since SOM performs neighborhood learning based on the similarity to the best matching unit, there is a problem that the same class is distributed and the boundary between classes becomes unclear. In this research, the learning rate of SOM is determined by weighted attribute information of the input data class. The proposed method reduced class dispersion about 50% and improves classification accuracy more than 1 % compared to conventional method by suppressing intervention in other classes.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Self-Organizing Map / Supervised Learning / Attribute Information / Handwriting Recognition
Paper # MBE2019-41,NC2019-32
Date of Issue 2019-10-04 (MBE, NC)

Conference Information
Committee MBE / NC
Conference Date 2019/10/11(2days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair Taishin Nomura(Osaka Univ.) / Hayaru Shouno(UEC)
Vice Chair Takashi Watanabe(Tohoku Univ.) / Kazuyuki Samejima(Tamagawa Univ)
Secretary Takashi Watanabe(Kyushu Univ.) / Kazuyuki Samejima(NAIST)
Assistant Yasuyuki Suzuki(Osaka Univ.) / Akihiro Karashima(Tohoku Inst. of Tech.) / Takashi Shinozaki(NICT) / Ken Takiyama(TUAT)

Paper Information
Registration To Technical Committee on ME and Bio Cybernetics / Technical Committee on Neurocomputing
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) An Optimization for Classification by Self-Organizing Maps Based on Attribute Information
Sub Title (in English)
Keyword(1) Self-Organizing Map
Keyword(2) Supervised Learning
Keyword(3) Attribute Information
Keyword(4) Handwriting Recognition
1st Author's Name Tetsuya Sato
1st Author's Affiliation Nihon University(Nihon Univ.)
2nd Author's Name Kazuma Tsuchida
2nd Author's Affiliation STUDIO ONE OR EIGHT(STUDIO ONE OR EIGHT)
3rd Author's Name Yukari Yamauti
3rd Author's Affiliation Nihon University(Nihon Univ.)
Date 2019-10-12
Paper # MBE2019-41,NC2019-32
Volume (vol) vol.119
Number (no) MBE-224,NC-225
Page pp.pp.59-63(MBE), pp.59-63(NC),
#Pages 5
Date of Issue 2019-10-04 (MBE, NC)