Advanced Data Mining

Scientifics responsibles: 
Pierre Comon (Gipsa-lab) , Massih-Reza Amini (LIG)

The goal of this research action is to propose new models, algorithms and tools for mining knowledge from the physical world, seen as a planetary sensor network producing distributed flows of digital data. The focus will be on online mining of distributed data flows that are digital traces of events, objects, usages or phenomena arising in the physical world. It is crucial to provide automatic tools to analyze those data flows in order to discover relevant information in them, and to do it online so that we can act on the physical world and take appropriate decisions if necessary.

The main challenges that will be addressed are related to dynamicity, multi-modality and decentralization:

  • Online mining of data streams

  • Mining multi-modal data

  • Decentralized data mining

Scientific committee: 
Pierre-Olivier Amblard (Gipsa-Lab), Renaud Blanch (LIG), Mauro Dalla Mura (Gipsa-lab), Ahlame Douzal (LIG), Olivier François (TIMC), Anne Guerin (Gipsa-lab), Pierre-Yves Guméry (TIMC), Nabil Layaida (Inria/LIG), Frédéric Pétrot (TIMA), Georges Quénot (LIG), Bertrand Rivet (Gipsa-lab), Michele Rombaut (Gipsa-lab), Jakob Verbeek (Inria).