Scientific projects

• Stochastic subsampling for factorizing huge matrices
• Fast independent component analysis solver for real data
• Convolutional Dictionary Learning Model at scales
• Learning from heterogeneous data
• Fast clustering with the Recursive Nearest agglomeration algorithm
• Magneto-encelography source localization
• Untangling high dimensional statistical inference problems using clustering and ensembling
• Causal inference with mediation using machine learning in high-dimensional settings
• Feature grouping as a Stochastic Regularizer for High-Dimensional Structured Data 
  

Partners and related efforts

• Paris-Saclay Center for Data Science – Phase 2: Parietal is the host of the data-scientists for the CDS
• Dataia: Parietal is funded by DataIA: MissingBigData project. Bertrand Thirion is the head of Dataia since January 2019.