éQuipe AppRentissage et MultimediA

Responsable : Thierry ARTIERES
QARMA devotes its research to machine learning and its applications to multimedia mining problems. It is particularly interested in theoretical aspects of machine learning, such as establishing consistency results for learning procedures, learning stochastic rational languages, dealing with non-IID data, developing new methods for operator-valued kernels and taking advantage of the information of multiview data. The problems tackled by the team stem from real-world problems from multimedia mining. It therefore entails the need for proper modeling/formalization, for scalable learning algorithms and for an awareness of recent developments in Big Data processing.

Web site:

Voir aussi dans «Calcul»