6 papers
Semi-knockoffs: a model-agnostic conditional independence testing method with finite-sample guarantees
Angel Reyero-Lobo, Bertrand Thirion, Pierre Neuvial
Conditional independence testing (CIT) is essential for reliable scientific discovery. It prevents spurious findings and enables controlled feature selection. Recent CIT methods ha…
Conditional Feature Importance revisited: Double Robustness, Efficiency and Inference
Angel Reyero-Lobo, Pierre Neuvial, Bertrand Thirion
Conditional Feature Importance (CFI) is a classical variable importance measure that accounts for the relationship between the studied feature and the others. However, CFI has not…
Cluster Size Matters: A Comparative Study of Notip and pARI for Post Hoc Inference in fMRI
Nils Peyrouset, Pierre Neuvial, Bertrand Thirion
All Resolutions Inference (ARI) is a post hoc inference method for functional Magnetic Resonance Imaging (fMRI) data analysis that provides valid lower bounds on the proportion of…
A principled approach for comparing Variable Importance
Angel Reyero-Lobo, Pierre Neuvial, Bertrand Thirion
Variable importance measures (VIMs) aim to quantify the contribution of each input covariate to the predictability of a given output. With the growing interest in explainable AI, n…
Inference post region selection
Dominique Bontemps, François Bachoc, Pierre Neuvial
Post-selection inference consists in providing statistical guarantees, based on a data set, that are robust to a prior model selection step on the same data set. In this paper, we…
When Knockoffs fail: diagnosing and fixing non-exchangeability of Knockoffs
Alexandre Blain, Angel Reyero Lobo, Julia Linhart +2
Knockoffs are a popular statistical framework that addresses the challenging problem of conditional variable selection in high-dimensional settings with statistical control. Such s…