{"id":"arxiv/2610.41442","url":"https://arxiv.org/abs/2610.41442","tags":["math.ST"],"score":0,"title":"Simultaneous inference for additive regression","domain":"math.ST","source":"arxiv","authors":["Donghyun Kim","Changwon Choi"],"reviews":0,"subject":"math.ST","summary":"This paper develops methods for simultaneous inference on component functions in additive models. We first propose bias-corrected smooth backfitting estimators inspired by least squares linear regression and establish their asymptotic properties. We then show that the distributions of the uniform errors of these estimators converge to a Gumbel-type distribution. Based on this distributional approximation, we construct simultaneous confidence bands and develop a practical residual-based procedure for testing the additive components. Simulation studies and empirical data analyses demonstrate…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-10-30","reactions":{}}