Advertisement

Maximal predictive clustering with order constraint: a linear and optimal algorithm

  • Laurent Guéguen
  • Régine Vignes
  • Jacques Lebbe
Part of the Studies in Classification, Data Analysis, and Knowledge Organization book series (STUDIES CLASS)

Abstract

According to the idea of the predictive power of a class in a classification (Gilmour 1951), our aim is to build maximal predictive classifications (Gower 1974) of objects in a sequence, that respect the total order defined by this sequence. We propose a new dynamic programming algorithm able to discover a maximal predictive partition and which complexity is linear with the length of the sequence and with the number of possible predictors. This algorithm accepts vast range of predictor shapes and may be used for numerous possible applications. We present an experiment of this clustering algorithm on biological sequences.

Keywords

Optimal clustering Maximal predictive classification Maximal predictive partition Constrained classification Biological sequences 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Archie, J. (1984). ‘A new look at the predictive value of numerical classifications’. Systematic Zoology 33, 30–51.CrossRefGoogle Scholar
  2. Colless, D. (1986). Classification et information. In P. Tassy (Ed.). ‘L’ordre et la diversité du vivant’. Fayard, pp. 121–142.Google Scholar
  3. Fisher, W. (1958). ‘On grouping for maximal homogeneity’. Journal of the American Statistical Association 53, 789–798.CrossRefGoogle Scholar
  4. Gilmour, J. (1951). ‘The development of taxonomic theory since 1851’. Nature 168, 400–402.CrossRefGoogle Scholar
  5. Gordon, A. (1981). Classification. Chapman and Hall.Google Scholar
  6. Gower, J. (1974). ‘Maximal predictive classification’. Biometrics 30, 643–654.CrossRefGoogle Scholar
  7. Lebbe, J. & R. Vignes (1994a). Classification maximalement prédictive sous contrainte d’ordre total. In: Actes de la société française de classification. Tours, pp. 127–128.Google Scholar
  8. Lebbe, J. & R. Vignes (1994b). Optimal maximal predictive constrained classification. In: Actes des septièmes entretiens du Centre Jacques Cartier. Lyon. pp. 81–83.Google Scholar
  9. Michalsky, R. & R. E. Stepp (1983). Automated construction of classifications: Conceptual clustering versus numerical taxonomy. In: IEEE Transactions on Pattern Analysis and Machine Intelligence. Vol PAMI-5 pp.396–410.Google Scholar
  10. Tassy, P. (1991). L’arbre à remonter le temps. Christian Bourgois.Google Scholar

Copyright information

© Springer-Verlag Berlin · Heidelberg 1998

Authors and Affiliations

  • Laurent Guéguen
    • 1
  • Régine Vignes
    • 1
  • Jacques Lebbe
    • 1
  1. 1.LIP6-Pôle IA UPMC-Boîte169Paris Cedex 05France

Personalised recommendations