A0939
Title: Communication-efficient distributed functional regression
Authors: Heng Lian - City university of Hong kong (Hong Kong) [presenting]
Abstract: The distributed estimation is considered via the communication-efficient surrogate loss approach of prior studies for functional linear regression in a unified framework, including functional generalized linear regression and functional quantile linear regression. The study of distributed functional regression is a methodologically important and technically non-trivial extension of the method to estimation in an infinite-dimensional Hilbert space. Statistical rates of the distributed estimator that match the rates of the standard estimator are established using all data.
