SELECT Research team
Model selection in statistical learning
- Leader : Pascal Massart
- Type : Project team
- Research center(s) : Saclay
- Field : Applied Mathematics, Computation and Simulation
- Theme : Optimization, machine learning and statistical methods
- Partner(s) : Université Paris-Sud (Paris 11),CNRS
- Collaborator(s) : Laboratoire de mathématiques d'Orsay de l'Université de Paris-Sud (LMO) (UMR8628)
SELECT is a project-team involved in statistical modelling. It is essentially concerned with model selection problems in statistical learning. SELECT is interested in hidden structure models, statistical pattern recognition, and most generally in statistical decision problems. Its application areas are Reliability, Microarray data analysis and Philogeny.
SELECT is aiming to propose tools in models and variables selection. Those tools are based on penalized likelihood criteria essentially designed from a non asymptotic point of view using concentration inequalities or from a Bayesian point of view taking profit of the modelling purpose. The research is oriented towards applications for supervised classification models and detection of abrupt changes in signal processing.
International and industrial relations
SELECT has regular collaborations with the Rearch Departement of EDF. Mambers of SELECT are in the European Network of Excellence PASCAL project (FP6).
Research teams of the same theme :
- BONUS - Big Optimization aNd UncertaintieS
- GEOSTAT - Geometry and Statistics in acquisition data
- INOCS - Integrated Optimization with Complex Structure
- MISTIS - Modelling and Inference of Complex and Structured Stochastic Systems
- MODAL - MOdel for Data Analysis and Learning
- RANDOPT - Randomized Optimisation
- REALOPT - Reformulations based algorithms for Combinatorial Optimization
- SEQUEL - Sequential Learning
- SIERRA - Statistical Machine Learning and Parsimony
- TAU - TAckling the Underspecified