Doubly stable feature selection perturbs the design matrix with increasing additive noise, fits a base selector like Lasso on each perturbed version, and aggregates selection frequencies to identify features stable across both subsampling and design noise levels.
Proceedings of the 26th Annual International Conference on Machine Learning , pages =
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
2D Stability Selection: Design Jittering for Doubly Stable Feature Selection
Doubly stable feature selection perturbs the design matrix with increasing additive noise, fits a base selector like Lasso on each perturbed version, and aggregates selection frequencies to identify features stable across both subsampling and design noise levels.