Machine learning and massive data analysis

This division deals with inference, predictive modeling and sequential optimization from complex data such as time series, functional data, network data

Scientific referents

Coordinator : Nicolas Vayatis

Scientific leaders :

Laurent Oudre
Mathilde Mougeot

Key figures

4 post-doc -- 17 phd students -- 3 internship students

Presentation of the thematic area

The expertise of the researchers at the Centre Borelli in machine learning algorithms and complex data modeling offers solid perspectives in academic and industrial projects where decisions must be made on the basis of data.

Starting from the double observation that the expert of the phenomenon cannot dispense with a statistical support to fully analyze a phenomenon and that the knowledge of the "physics" of the phenomenon is not replaceable, the statistical learning methods are limited without the contribution of an expert.

Thus, the researchers of the Centre Borelli are generally immersed in interdisciplinary teams, in companies or in laboratories, and develop algorithms in close collaboration with the end users.

Key words

Statistical and machine learning; Neural networks.

Topics covered

Scoring, ranking, theory and algorithms
Inference and prediction for very large graphs
Sequential design of experiments and active learning
Machine learning methods for multivariate time series

Key facts

Creation of the Chair IDAML

Applications

Reconnaissance de formes
E-marketing,
Energy
Finance and economy
Health and biomedical field
Sensor networks