Open Access
Translator Disclaimer
2011 Subsampling algorithms for semidefinite programming
Alexandre d’Aspremont
Stoch. Syst. 1(2): 274-305 (2011). DOI: 10.1214/10-SSY018

Abstract

We derive a stochastic gradient algorithm for semidefinite optimization using randomization techniques. The algorithm uses subsampling to reduce the computational cost of each iteration and the subsampling ratio explicitly controls granularity, i.e. the tradeoff between cost per iteration and total number of iterations. Furthermore, the total computational cost is directly proportional to the complexity (i.e. rank) of the solution. We study numerical performance on some large-scale problems arising in statistical learning.

Citation

Download Citation

Alexandre d’Aspremont. "Subsampling algorithms for semidefinite programming." Stoch. Syst. 1 (2) 274 - 305, 2011. https://doi.org/10.1214/10-SSY018

Information

Published: 2011
First available in Project Euclid: 24 February 2014

zbMATH: 1291.90169
MathSciNet: MR2949542
Digital Object Identifier: 10.1214/10-SSY018

Subjects:
Primary: 90C15, 90C22

Rights: Copyright © 2011 INFORMS Applied Probability Society

JOURNAL ARTICLE
32 PAGES


SHARE
Vol.1 • No. 2 • 2011
Back to Top