Latent Block Models

Course Code : IA344

Time Hours : 18 hours

Time Periods : not available yet

Lecturer : Mohamed Nadif, Professor, mohamed.nadif [at] parisdescartes.fr

Objective

The goal is to treat the problem of co-clustering. Conventional techniques are described for solving current problems from different fields such as bioinformatics, text mining, web mining and marketing. New methods derived from the latent block models will be studied and illustrated with real data sets.

Acquired Skills

In which context we can use the co-clustering, the traditional methods and the new ones arising from the latent block models.

Contents

  • Clustering variables
  • Interest of co-clustering in different fieldsspeech processing
  • Nature of co-clusters
  • Criteria, algorithms and different approaches
  • Latent Block Models
  • Applications : binary, continuous, categorical and occurrences data
  • EM and Classification EM algorithms
  • Applications : text-mining, web mining and bioinformatics

References

  • Data Analysis, Gérard Govaert, ISTE Ltd and John Wiley & Sons Inc (5 août 2009)