A Graph Theoretic Criterion for Determining the Number of Clusters in a Data Set.

Krolak-Schwerdt, Sabine; Eckes, Thomas

Notes FAQ Contact Us

Back to results

Peer reviewed

ERIC Number: EJ464856

Record Type: Journal

Publication Date: 1992

Pages: N/A

Abstractor: N/A

ISBN: N/A

ISSN: ISSN-0027-3171

EISSN: N/A

Available Date: N/A

A Graph Theoretic Criterion for Determining the Number of Clusters in a Data Set.

Krolak-Schwerdt, Sabine; Eckes, Thomas

Multivariate Behavioral Research, v27 n4 p541-65 1992

Procedures for determining the number of clusters in a data set are explored. A proposed stopping rule, the GRAPH criterion, is compared to four stopping rules currently in use. The GRAPH criterion's mathematically attractive properties and utility in solving the number-of-clusters problem are demonstrated. (SLD)

Descriptors: Cluster Analysis, Data Collection, Equations (Mathematics), Evaluation Criteria, Mathematical Models

Publication Type: Reports - Evaluative; Journal Articles

Education Level: N/A

Audience: N/A

Language: English

Sponsor: N/A

Authoring Institution: N/A

Grant or Contract Numbers: N/A

Author Affiliations: N/A