A Generalized Unsupervised Competitive Learning Scheme

Ferdinand PEPER; Hideki NODA

A Generalized Unsupervised Competitive Learning Scheme

Ferdinand PEPER, Hideki NODA

Full Text Views

0

Share
Cite this

Summary :

In this article a Neural Network learning scheme is described, which is a generalization of VQ (Vector Quantization) and ART2a (a simplified version of Adaptive Resonance Theory 2). The basic differences between VQ and ART2a will be exhibited and it will be shown how these differences are covered by the generalized scheme. The generalized scheme enables a rich set of variations on VQ and ART2a. One such variation uses the expression ||I||²+||z_j||²/||z_j||sin(I,z_j), as the distance measure between input vector I and weight vector z_j. This variation tends to be more robust to noise than ART2a, as is shown by experiments we performed. These experiments use the same data-set as the ART2a experiments in Ref.(3).

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E76-A No.5 pp.834-841

Publication Date: 1993/05/25

Publicized

Online ISSN

DOI

Type of Manuscript: PAPER

Category: Neural Networks

Cite this

Copy

Ferdinand PEPER, Hideki NODA, "A Generalized Unsupervised Competitive Learning Scheme" in IEICE TRANSACTIONS on Fundamentals, vol. E76-A, no. 5, pp. 834-841, May 1993, doi: .
Abstract: In this article a Neural Network learning scheme is described, which is a generalization of VQ (Vector Quantization) and ART2a (a simplified version of Adaptive Resonance Theory 2). The basic differences between VQ and ART2a will be exhibited and it will be shown how these differences are covered by the generalized scheme. The generalized scheme enables a rich set of variations on VQ and ART2a. One such variation uses the expression ||I||²+||z_j||²/||z_j||sin(I,z_j), as the distance measure between input vector I and weight vector z_j. This variation tends to be more robust to noise than ART2a, as is shown by experiments we performed. These experiments use the same data-set as the ART2a experiments in Ref.(3).
URL: https://globals.ieice.org/en_transactions/fundamentals/10.1587/e76-a_5_834/_p

Copy

@ARTICLE{e76-a_5_834,
author={Ferdinand PEPER, Hideki NODA, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={A Generalized Unsupervised Competitive Learning Scheme},
year={1993},
volume={E76-A},
number={5},
pages={834-841},
abstract={In this article a Neural Network learning scheme is described, which is a generalization of VQ (Vector Quantization) and ART2a (a simplified version of Adaptive Resonance Theory 2). The basic differences between VQ and ART2a will be exhibited and it will be shown how these differences are covered by the generalized scheme. The generalized scheme enables a rich set of variations on VQ and ART2a. One such variation uses the expression ||I||²+||z_j||²/||z_j||sin(I,z_j), as the distance measure between input vector I and weight vector z_j. This variation tends to be more robust to noise than ART2a, as is shown by experiments we performed. These experiments use the same data-set as the ART2a experiments in Ref.(3).},
keywords={},
doi={},
ISSN={},
month={May},}

Copy

TY - JOUR
TI - A Generalized Unsupervised Competitive Learning Scheme
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 834
EP - 841
AU - Ferdinand PEPER
AU - Hideki NODA
PY - 1993
DO -
JO - IEICE TRANSACTIONS on Fundamentals
SN -
VL - E76-A
IS - 5
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - May 1993
AB - In this article a Neural Network learning scheme is described, which is a generalization of VQ (Vector Quantization) and ART2a (a simplified version of Adaptive Resonance Theory 2). The basic differences between VQ and ART2a will be exhibited and it will be shown how these differences are covered by the generalized scheme. The generalized scheme enables a rich set of variations on VQ and ART2a. One such variation uses the expression ||I||²+||z_j||²/||z_j||sin(I,z_j), as the distance measure between input vector I and weight vector z_j. This variation tends to be more robust to noise than ART2a, as is shown by experiments we performed. These experiments use the same data-set as the ART2a experiments in Ref.(3).
ER -