Preprint
Aug 2026
Generalization Error Estimation for Primal--Dual Algorithms in Non-Smooth Regression
A general recursive framework that includes the Chambolle--Pock algorithm and related primal--dual splitting methods is developed, which proves finite-sample guarantees for both estimators and establishes a matched-Gaussian universality result beyond Gaussian designs.
Kai Tan, Pierre C. Bellec
· 0 citations