On the low-rank approach for semidefinite programs arising in synchronization and community detection

Afonso Bandeira, Nicolas Boumal, Vladislav Voroninski

Research output: Contribution to journalConference article

Abstract

To address difficult optimization problems, convex relaxations based on semidefinite programming are now common place in many fields. Although solvable in polynomial time, large semidefinite programs tend to be computationally challenging. Over a decade ago, exploiting the fact that in many applications of interest the desired solutions are low rank, Burer and Monteiro proposed a heuristic to solve such semidefinite programs by restricting the search space to low-rank matrices. The accompanying theory does not explain the extent of the empirical success. We focus on Synchronization and Community Detection problems and provide theoretical guarantees shedding light on the remarkable efficiency of this heuristic.

Original languageEnglish (US)
Pages (from-to)361-382
Number of pages22
JournalJournal of Machine Learning Research
Volume49
Issue numberJune
StatePublished - Jun 6 2016
Event29th Conference on Learning Theory, COLT 2016 - New York, United States
Duration: Jun 23 2016Jun 26 2016

Fingerprint

Community Detection
Semidefinite Program
Synchronization
Polynomials
Heuristics
Low-rank Matrices
Convex Relaxation
Semidefinite Programming
Search Space
Polynomial time
Tend
Optimization Problem

Keywords

  • Burer-Monteiro heuristic
  • Community Detection
  • SDPLR
  • Semidefinite Programming
  • Synchronization

ASJC Scopus subject areas

  • Software
  • Control and Systems Engineering
  • Statistics and Probability
  • Artificial Intelligence

Cite this

On the low-rank approach for semidefinite programs arising in synchronization and community detection. / Bandeira, Afonso; Boumal, Nicolas; Voroninski, Vladislav.

In: Journal of Machine Learning Research, Vol. 49, No. June, 06.06.2016, p. 361-382.

Research output: Contribution to journalConference article

Bandeira, Afonso ; Boumal, Nicolas ; Voroninski, Vladislav. / On the low-rank approach for semidefinite programs arising in synchronization and community detection. In: Journal of Machine Learning Research. 2016 ; Vol. 49, No. June. pp. 361-382.
@article{de46d31b77c6475cb7b82f046928495a,
title = "On the low-rank approach for semidefinite programs arising in synchronization and community detection",
abstract = "To address difficult optimization problems, convex relaxations based on semidefinite programming are now common place in many fields. Although solvable in polynomial time, large semidefinite programs tend to be computationally challenging. Over a decade ago, exploiting the fact that in many applications of interest the desired solutions are low rank, Burer and Monteiro proposed a heuristic to solve such semidefinite programs by restricting the search space to low-rank matrices. The accompanying theory does not explain the extent of the empirical success. We focus on Synchronization and Community Detection problems and provide theoretical guarantees shedding light on the remarkable efficiency of this heuristic.",
keywords = "Burer-Monteiro heuristic, Community Detection, SDPLR, Semidefinite Programming, Synchronization",
author = "Afonso Bandeira and Nicolas Boumal and Vladislav Voroninski",
year = "2016",
month = "6",
day = "6",
language = "English (US)",
volume = "49",
pages = "361--382",
journal = "Journal of Machine Learning Research",
issn = "1532-4435",
publisher = "Microtome Publishing",
number = "June",

}

TY - JOUR

T1 - On the low-rank approach for semidefinite programs arising in synchronization and community detection

AU - Bandeira, Afonso

AU - Boumal, Nicolas

AU - Voroninski, Vladislav

PY - 2016/6/6

Y1 - 2016/6/6

N2 - To address difficult optimization problems, convex relaxations based on semidefinite programming are now common place in many fields. Although solvable in polynomial time, large semidefinite programs tend to be computationally challenging. Over a decade ago, exploiting the fact that in many applications of interest the desired solutions are low rank, Burer and Monteiro proposed a heuristic to solve such semidefinite programs by restricting the search space to low-rank matrices. The accompanying theory does not explain the extent of the empirical success. We focus on Synchronization and Community Detection problems and provide theoretical guarantees shedding light on the remarkable efficiency of this heuristic.

AB - To address difficult optimization problems, convex relaxations based on semidefinite programming are now common place in many fields. Although solvable in polynomial time, large semidefinite programs tend to be computationally challenging. Over a decade ago, exploiting the fact that in many applications of interest the desired solutions are low rank, Burer and Monteiro proposed a heuristic to solve such semidefinite programs by restricting the search space to low-rank matrices. The accompanying theory does not explain the extent of the empirical success. We focus on Synchronization and Community Detection problems and provide theoretical guarantees shedding light on the remarkable efficiency of this heuristic.

KW - Burer-Monteiro heuristic

KW - Community Detection

KW - SDPLR

KW - Semidefinite Programming

KW - Synchronization

UR - http://www.scopus.com/inward/record.url?scp=85061238935&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85061238935&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:85061238935

VL - 49

SP - 361

EP - 382

JO - Journal of Machine Learning Research

JF - Journal of Machine Learning Research

SN - 1532-4435

IS - June

ER -