Partially linear functional quantile regression in a reproducing kernel Hilbert space. Issue 4 (2nd October 2022)

• Record Type: Journal Article
• Authors: Zhou, Yan; Zhang, Weiping; Lin, Hongmei; Lian, Heng
• Is Part Of: Journal of nonparametric statistics. Volume 34:Issue 4(2022)
• Page Start: 789
• Page End: 803
• Publication Date: 2022-10-02
• Subjects: Convergence rate -- prediction risk -- quantile regression -- rademacher complexity; 62G06; Nonparametric statistics -- Periodicals; 519.5
• DOI: 10.1080/10485252.2022.2073354
• Languages: English
• ISSNs: 1048-5252

Abstract

Abstract : We consider quantile functional regression with a functional part and a scalar linear part. We establish the optimal prediction rate for the model under mild assumptions in the reproducing kernel Hilbert space (RKHS) framework. Under stronger assumptions related to the capacity of the RKHS, the non-functional linear part is shown to have asymptotic normality. The estimators are illustrated in simulation studies.