INTERPRETING THE KOHONEN SELF-ORGANIZING FEATURE MAP USING CONTIGUITY-CONSTRAINED CLUSTERING

被引：77

作者：

MURTAGH, F ^{[1
]}

机构：

[1] EUROPEAN SPACE AGCY,DEPT SPACE SCI,DIV ASTROPHYS,F-75738 PARIS 15,FRANCE

来源：

PATTERN RECOGNITION LETTERS | 1995年 / 16卷 / 04期

关键词：

CLUSTER ANALYSIS; NEURAL NETWORKS; HIERARCHICAL CLUSTERING; REGIONALIZATION; SEGMENTATION; DATA ANALYSIS; IMAGE PROCESSING;

D O I：

10.1016/0167-8655(94)00113-H

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

An interpretation phase is proposed, to complement usage of the Kohonen self-organizing feature map (SOFM) method. This segments the SOFM output, using an agglomerative contiguity-constrained clustering method. We discuss why such a clustering method, which respects contiguity information, should be used. The contiguity-constrained clustering method implemented uses array operations as far as possible, and is thus quite efficient. An application to an astronomical catalog of about a quarter million objects is presented.

引用

页码：399 / 408

页数：10

共 24 条

[1] HIERARCHY IN PICTURE SEGMENTATION - A STEPWISE OPTIMIZATION APPROACH [J].

BEAULIEU, JM ;

GOLDBERG, M .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1989, 11 (02) :150-163

[2]

BOLLER T, 1992, ASTRON ASTROPHYS, V259, P101

[3]

Gordon A. D., 1981, CLASSIFICATION METHO

[4]

Hartigan JohnA., 1975, CLUSTERING ALGORITHM

[5] LOCALLY SENSITIVE METHOD FOR CLUSTER-ANALYSIS [J].

KITTLER, J .

PATTERN RECOGNITION, 1976, 8 (01) :23-33

[6] SINGLE-LINK CHARACTERISTICS OF A MODE-SEEKING CLUSTERING ALGORITHM - COMMENTS [J].

KITTLER, J .

PATTERN RECOGNITION, 1979, 11 (01) :71-73

[7] THE SELF-ORGANIZING MAP [J].

KOHONEN, T .

PROCEEDINGS OF THE IEEE, 1990, 78 (09) :1464-1480

[8]

KOHONEN T, 1992, SELF ORG MAP PROGRAM

[9]

KOHONEN T, 1988, P CONNECTIONISM PERS

[10]

MEURS EJA, 1988, ASTRON ASTROPHYS, V206, P53

← 1 2 3 →