ADVANSE: ADVanced Analytics for data SciencE
The ADVANSE (Advanced Analytics for Data SciencE) team was created in 2014 and is attached to the LIRMM laboratory. This team is directed by Pascal Poncelet. The research activities carried out by the ADVANSE team are part of the field of large database analysis in order to extract new knowledge. They concern the following three axes:
- Data mining (DM) which pursues historical work with particular emphasis on pattern mining approaches integrating either new dimensions such as, for example, the spatial dimension in the case of spatio-temporal patterns and trajectories or the dimensions associated with the different types of arcs in multigraphs;
- Analytical Visualization (VA) that emphasizes analytical reasoning facilitated by interactive visual interfaces to help decision makers understand their data or better understand the knowledge extracted from these data;
- Machine learning (ML) with definition of new approaches (e.g. detection of rare clusters, topics labeling), and special emphases on small or very large datasets via traditional approaches (e.g. SVM, Gradient Boosting, active learning) or more recent ones (e.g. deep learning). To cope with large numbers of dimensions (e.g. images, sounds, texts), deep learning has shown to be very effective for classification. However, the explainability of the results, or the definition of a efficient architecture remain very experimental exercises and constitute a real challenge for the scientific community.
The ADVANSE team, with this three axes, develops works on theoretical as well as experimental bases to tackle the associated problems.Links: