IS2 has been dissolved on 09/01/2003
It has been replaced by MISTIS
Project-Team Presentation
This project aims to research into statistical modelling for industrial and medical field. It deals with diagnosis, reliability, industrial failure time analysis, models for hospital lengths of stay, statistical image analysis and finance. The chosen methods involve generalized linear models, hidden structure models identified by stochastic algorithms, adaptative estimation, conditionally heterocedastic models for time series, dynamical systems identification.
Research themes
- Statistical modelling from a maximum entropy principle.
- Analysis of stochastic relaxation algorithms.
- Models of public health, hospital activity, nosocomial infections.
- Statistical biomedical or industrial diagnosis.
- Analysis of failure time of complex systems ruled by Weibull distributions.
- Robust methods and rare events data.
- Selecting model methods.
- Heterocesdatic models.
- Random effects models.
- Non parametric modelling.
International and industrial relations
- Collaboration with the Timc laboratory at Imag for all medical research.
- Contracts with Experimental Feedback and Reliability laboratories of the EDF-DER and the CEA/Grenoble dosimetry unit.
- Collaboration with the department of Statistics of the Washington University (Seattle) and Humboldt University in Berlin.
Scientific leader
Gilles CELEUX
+33 4 76 61 53 25
Gilles.Celeux@inria.fr
Secretary : +33 4 76 61 53 63
