MAM is a bilevel-optimized sparse additive model that auto-learns loss weights via an MLP to handle complex noise, with convergence and variable selection consistency guarantees.
Tibshirani, Regression shrinkage and selection via the lasso, Journal of the Royal Statistical Society, Series B 73 (3) (1994) 267–288
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Meta Additive Model: Interpretable Sparse Learning With Auto Weighting
MAM is a bilevel-optimized sparse additive model that auto-learns loss weights via an MLP to handle complex noise, with convergence and variable selection consistency guarantees.