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


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


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

Computer Science > Machine Learning

Title: Riemannian stochastic variance reduced gradient on Grassmann manifold

Abstract: Stochastic variance reduction algorithms have recently become popular for minimizing the average of a large, but finite, number of loss functions. In this paper, we propose a novel Riemannian extension of the Euclidean stochastic variance reduced gradient algorithm (R-SVRG) to a compact manifold search space. To this end, we show the developments on the Grassmann manifold. The key challenges of averaging, addition, and subtraction of multiple gradients are addressed with notions like logarithm mapping and parallel translation of vectors on the Grassmann manifold. We present a global convergence analysis of the proposed algorithm with decay step-sizes and a local convergence rate analysis under fixed step-size with some natural assumptions. The proposed algorithm is applied on a number of problems on the Grassmann manifold like principal components analysis, low-rank matrix completion, and the Karcher mean computation. In all these cases, the proposed algorithm outperforms the standard Riemannian stochastic gradient descent algorithm.
Subjects: Machine Learning (cs.LG); Numerical Analysis (math.NA); Optimization and Control (math.OC); Machine Learning (stat.ML)
Cite as: arXiv:1605.07367 [cs.LG]
  (or arXiv:1605.07367v3 [cs.LG] for this version)

Submission history

From: Hiroyuki Kasai [view email]
[v1] Tue, 24 May 2016 10:36:32 GMT (4908kb)
[v2] Thu, 26 May 2016 12:55:11 GMT (4908kb)
[v3] Sun, 9 Apr 2017 06:17:52 GMT (1707kb,D)

Link back to: arXiv, form interface, contact.