Variable Importance in High-Dimensional Settings Requires Grouping

1 code implementation • 18 Dec 2023 • Ahmad Chamma, Bertrand Thirion, Denis A. Engemann

Furthermore, as handling groups with high cardinality (such as a set of observations of a given modality) are both time-consuming and resource-intensive, we also introduce a new stacking approach extending the DNN architecture with sub-linear layers adapted to the group structure.

Manifold-regression to predict from MEG/EEG brain signals without source modeling

1 code implementation • NeurIPS 2019 • David Sabbagh, Pierre Ablin, Gael Varoquaux, Alexandre Gramfort, Denis A. Engemann

We show that Wasserstein and geometric distances allow perfect out-of-sample prediction on the generative models.

EEG regression +1