Research themes
-
Investigation of multi-dimensional datawarehouses able to store complex and heterogeneous data. Extension of query languages with data mining functionalities for a more relevant and access to relevant information.
-
Investigation of automatic model and knowledge acquisition methods from massive data with a focus on temporal and spatial patterns. Data stream analysis and model updating in order take into account data novelty. Investigation of incremental learning methods.
-
Development of decision-aiding tools to help a user analyze data at hand (simulation results, for instance) by extracting interesting patterns and recommending actions. Investigation of interactive learning methods tools.
-
Investigation of the interaction between diagnosis and decision in uncertain universes and, particularly, recommending actions from simulation results.
For more details see the annual project activity report.
|