We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

math.OC

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Optimization and Control

Title: Nearly Optimal Stochastic Approximation for Online Principal Subspace Estimation

Abstract: Principal component analysis (PCA) has been widely used in analyzing high-dimensional data. It converts a set of observed data points of possibly correlated variables into a set of linearly uncorrelated variables via an orthogonal transformation. To handle streaming data and reduce the complexities of PCA, (subspace) online PCA iterations were proposed to iteratively update the orthogonal transformation by taking one observed data point at a time. Existing works on the convergence of (subspace) online PCA iterations mostly focus on the case where sample are almost surely uniformly bounded. In this paper, we analyze the convergence of a subspace online PCA iteration under more practical assumption and obtain a nearly optimal finite-sample error bound. Our convergence rate almost matches the minimax information lower bound. We prove that the convergence is nearly global in the sense that the subspace online PCA iteration is convergent with high probability for random initial guesses. This work also leads to a simpler proof of the recent work on analyzing online PCA for the first principal component only.
Comments: 34 pages. Accepted for publication in SCIENCE CHINA Mathematics
Subjects: Optimization and Control (math.OC)
MSC classes: 65F99, 62H25, 68W27, secondary 62H12
Cite as: arXiv:1711.06644 [math.OC]
  (or arXiv:1711.06644v3 [math.OC] for this version)

Submission history

From: Xin Liang [view email]
[v1] Fri, 17 Nov 2017 17:42:36 GMT (46kb)
[v2] Wed, 15 Aug 2018 00:03:15 GMT (50kb)
[v3] Tue, 24 May 2022 08:29:35 GMT (56kb)

Link back to: arXiv, form interface, contact.