Frequency sensitive competitive learning for clustering on high-dimensional hyperspheres

被引：16

作者：

Banerjee, A ^{[1
]}

Ghosh, J ^{[1
]}

机构：

[1] Univ Texas, Dept Elect & Comp Engn, Austin, TX 78712 USA

来源：

PROCEEDING OF THE 2002 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS, VOLS 1-3 | 2002年

关键词：

D O I：

10.1109/IJCNN.2002.1007755

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

This paper derives three competitive learning mechanisms from first principles to obtain clusters of comparable sizes when both inputs and representatives are normalized. These mechanisms are very effective in achieving balanced grouping of inputs in high dimensional spaces, as illustrated by experimental results on clustering two popular text data sets in 26,099 and 21,839 dimensional spaces respectively.

引用

页码：1590 / 1595

页数：4

共 21 条

[11] ADAPTIVE PATTERN-CLASSIFICATION AND UNIVERSAL RECODING .1. PARALLEL DEVELOPMENT AND CODING OF NEURAL FEATURE DETECTORS [J].

GROSSBERG, S .

BIOLOGICAL CYBERNETICS, 1976, 23 (03) :121-134

[12] Data clustering: A review [J].

Jain, AK ;

Murty, MN ;

Flynn, PJ .

ACM COMPUTING SURVEYS, 1999, 31 (03) :264-323

[13]

MARDIA KV, 1975, J R STAT SOC B, V37, P349

[14]

McLachlan Norman William, 1955, BESSEL FUNCTIONS ENG

[15]

Mitchell T., 1997, Machine Learning, V7, P2

[16]

RUMELHART DE, 1985, COGNITIVE SCI, V9, P75, DOI 10.1207/s15516709cog0901_5

[17] High-dimensional clustering using frequency sensitive competitive learning [J].

Scheunders, P ;

De Backer, S .

PATTERN RECOGNITION, 1999, 32 (02) :193-202

[18]

SINKKONEN J, 2001, NEURAL COMPUTATION

[19]

STREHL A, 2000, P 17 INT C HIGH PERF, P525

[20]

STREHL A, 2000, P SPIE C DAT MIN KNO, V2, P33

← 1 2 3 →