Competitive learning algorithms founded on adaptivity and sensitivity deletion methods

Authors
Maeda, M Miyajima, H
Citation
M. Maeda et H. Miyajima, Competitive learning algorithms founded on adaptivity and sensitivity deletion methods, IEICE T FUN, E83A(12), 2000, pp. 2770-2774
Citations number
10
Categorie Soggetti
Eletrical & Eletronics Engineeing
Journal title
IEICE TRANSACTIONS ON FUNDAMENTALS OF ELECTRONICS COMMUNICATIONS AND COMPUTER SCIENCES
ISSN journal
09168508 → ACNP
Volume
E83A
Issue
12
Year of publication
2000
Pages
2770 - 2774
Database
ISI
SICI code
0916-8508(200012)E83A:12<2770:CLAFOA>2.0.ZU;2-A
Abstract
This paper describes two competitive learning algorithms from the viewpoint of deleting mechanisms of weight (reference) vectors. The techniques are t ermed the adaptivity and sensitivity deletions participated in the criteria of partition error and distortion error, respectively. Experimental result s show the effectiveness of the proposed approaches in the average age dist ortion.