Skip to main content

Distributed Estimation for Principal Component Analysis: An Enlarged Eigenspace Analysis

Author(s): Chen, Xi; Lee, Jason D; Li, He; Yang, Yun

To refer to this page use:
Full metadata record
DC FieldValueLanguage
dc.contributor.authorChen, Xi-
dc.contributor.authorLee, Jason D-
dc.contributor.authorLi, He-
dc.contributor.authorYang, Yun-
dc.identifier.citationChen, Xi, Lee, Jason D, Li, He, Yang, Yun. (2022). Distributed Estimation for Principal Component Analysis: An Enlarged Eigenspace Analysis. Journal of the American Statistical Association, 117 (540), 1775 - 1786. doi:10.1080/01621459.2021.1886937en_US
dc.description.abstractThe growing size of modern datasets brings many challenges to the existing statistical estimation approaches, which calls for new distributed methodologies. This article studies distributed estimation for a fundamental statistical machine learning problem, principal component analysis (PCA). Despite the massive literature on top eigenvector estimation, much less is presented for the top-L-dim (L > 1) eigenspace estimation, especially in a distributed manner. We propose a novel multi-round algorithm for constructing top-L-dim eigenspace for distributed data. Our algorithm takes advantage of shift-and-invert preconditioning and convex optimization. Our estimator is communication-efficient and achieves a fast convergence rate. In contrast to the existing divide-and-conquer algorithm, our approach has no restriction on the number of machines. Theoretically, the traditional Davis–Kahan theorem requires the explicit eigengap assumption to estimate the top-L-dim eigenspace. To abandon this eigengap assumption, we consider a new route in our analysis: instead of exactly identifying the top-L-dim eigenspace, we show that our estimator is able to cover the targeted top-L-dim population eigenspace. Our distributed algorithm can be applied to a wide range of statistical problems based on PCA, such as principal component regression and single index model. Finally, we provide simulation studies to demonstrate the performance of the proposed distributed estimator.en_US
dc.format.extent1775 - 1786en_US
dc.relation.ispartofJournal of the American Statistical Associationen_US
dc.rightsAuthor's manuscripten_US
dc.titleDistributed Estimation for Principal Component Analysis: An Enlarged Eigenspace Analysisen_US
dc.typeJournal Articleen_US

Files in This Item:
File Description SizeFormat 
2004.02336.pdf5.01 MBAdobe PDFView/Download

Items in OAR@Princeton are protected by copyright, with all rights reserved, unless otherwise indicated.