Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.480835
Title: An evaluation of cluster analysis and related multivariate techniques for operational research
Author: Newport, Ray
Awarding Body: City University London
Current Institution: City, University of London
Date of Award: 1974
Availability of Full Text:
Access through EThOS:
Access through Institution:
Abstract:
The following work is an investigation into methods of Cluster Analysis and Ordination. The main objective of this thesis has been to investigate the capabilities of these methods for practical usage. An important subsidiary aim has been to collect together related work which has been carried out in many different areas - ecology, biology, archaeology, psychology, etc., - into one work. After a brief introduction to the general concepts of multivariate analysis in Section A of the thesis, Section B gives an introductory account of the methods of Clustering, Ordination and Seriation, putting them into the context of the by now better established multivariate techniques. Section C considers Cluster Analysis in depth, explaining and examining various methods reported in the literature, together with methods developed by the author. The suitability of the methods for practical use is discussed and, decision rules are set out for the choice of method to be used in any particular study, based on the results of extensive comparative tests of the methods. In Section D the various ordination methods are. considered, giving an, overall viere and relating the methods to each other. Particular emphasis is paid to the rather neglected metric methods. Section E, after a survey of published applications of the methods, suggests new areas where the methods previously discussed could be valuable aids for data investigation and problem solving. An Addenda is included which describes several operational research case studies using these methods. Computer programs are given for the most successful of the newly introduced cluster methods, and an extensive reference section is also included.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.480835  DOI: Not available
Keywords: HB Economic Theory Mathematical statistics Operations research
Share: