The paper by Lei Shi and his collaborator Zhengchu Guo, entitled "Optimal rates for coefficient-based regularized regression", has been published in the journal "Appl. Comput. Harmon. Anal." in 2019.
In this paper, they study distributed learning with multi-penalty regularization based on a divide-and-conquer approach. Using Neumann expansion and a second order decomposition on difference of operator inverses approach, optimal learning rates for distributed multi-penalty regularization was derived. As a byproduct, they also deduce optimal learning rates for distributed manifold learning, which was not given in the previous literature.
220 Handan Rd., Yangpu District, Shanghai （ 200433 ） | Operator：+86 21 65642222
Copyright © 2016 FUDAN University. All Rights Reserved