Develops gradient descent and variational inequality algorithms for simultaneous estimation of generalized additive index models, with unified convergence guarantees and reported empirical gains over stage-wise methods.
Beyond maximum likelihood: Variational inequality estimation for generalized linear models.arXiv preprint arXiv:2511.03087, 2025
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
-
Efficient First-Order Methods for Estimating Generalized Additive Index Models
Develops gradient descent and variational inequality algorithms for simultaneous estimation of generalized additive index models, with unified convergence guarantees and reported empirical gains over stage-wise methods.