Advertisement

Pertinence for a Classification

  • N. Nicoloyannis
  • M. Terrenoire
  • D. Tounissoux
Conference paper
Part of the Studies in Classification, Data Analysis, and Knowledge Organization book series (STUDIES CLASS)

Abstract

The Clique Partitioning Problem constitutes a general framework for clustering, when similarities assume positive and negative values. The effectiveness of various heuristic methods of solving the above problem has been proved by several authors.

In order to evaluate the validity of one optimal classification P* obtained by this approach, we propose a hierarchical process, working on the clusters of P*. Thus, we obtain a family of classifications, and the concept of pertinence enables us to detect the interesting classifications.

Key words

Clique Partitioning Simulated annealing Tabu search Pertinence 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    De Amorim S. & Barthélémy & J.P., Ribeiro C. (1992). Clustering and Clique Partitioning: Simulated Annealing and Tabu Search Approaches, in Journal of Classification, 9, 17–41.Google Scholar
  2. [2]
    Bock H. (1994). Classification and Clustering: Problems for the Future, in: E. Diday et al (eds. New Approaches in Classification and Data Analysis, Springer Verlag, 3–24.Google Scholar
  3. [3]
    Grotschel M. & Wakabayashi Y. (1989) A Cutting Plane Algorithm for a Clustering Problem, in: Mathematical Programming, 45, 59–96.Google Scholar
  4. [4]
    Hansen P. & Jaumard B. & Sanlaville E. (1994). Partitionning Problems in Cluster Analysis: a Review of Mathematical Programming Approaches in: New Approaches in Classification and Data Analysis, E. Diday et al (eds), Springer Verlag, 228–240.Google Scholar
  5. [5]
    Marcotorchino J.F. & Michaud P. (1981). Heuristic Approach to the Similarity Aggregation Problem, in: Methods of Operations Research, 43, 395–404.Google Scholar
  6. [6]
    Saporta G. (1990). Probabilités, Analyse des Données et Statistiques, Technip, ParisGoogle Scholar

Copyright information

© Springer-Verlag Berlin · Heidelberg 1998

Authors and Affiliations

  • N. Nicoloyannis
    • 1
  • M. Terrenoire
    • 1
  • D. Tounissoux
    • 1
  1. 1.ESA 5047. Méthodes d’Analyse des Systèmes et des StructuresUniversité Claude-Bernard Lyon 1. Bât. 101Villeurbanne CedexFrance

Personalised recommendations