Project overview

IS2 is a project of the research unit Rhône-Alpes of INRIA (INRIA: National Institute of Research in Computer Science and Robotic). The project is concerned with statistical modelling. Emphasis is placed on incomplete data models. Its main areas of applications are biomedical statistic and failure time models. In 1996, IS2 developed activities statistical modelling through a maximum entropy principle, generalized linear models with random effects, stochastic algorithms, Bayesian statistical analysis of industrial failure times, competing risk models, tails of distributions, and medical applications (diagnosis of haemorragic infarct, models for thrombose's increase, and analysis of hospital length of stays via a mixture of exponential distributions).

Keywords: Statistical modeling, incomplete data, heteroscedasticity, generalized linear models, maximum entropy method, stochastics algorithms, diagnosis aid, survival analysis, rare events.

For further information please contact:

Gilles Celeux (

INRIA Rhône-Alpes, 655 avenue de l'Europe, 38330 Montbonnot Saint-Martin, France Telephone: (33) 0476 61 53 55 , Telefax: +(33) 04 76 61 52 07.

Last update 1-Oct-98; This web page is maintained by and


Back to the IS2 Home Page.