Local minima and convergence in low-rank semidefinite programming

被引:249
作者
Burer, S [1 ]
Monteiro, RDC
机构
[1] Univ Iowa, Dept Management Sci, Iowa City, IA 52242 USA
[2] Georgia Inst Technol, Sch Ind & Syst Engn, Atlanta, GA 30332 USA
关键词
semidefinite programming; low-rank matrices; vector programming; combinatorial optimization; nonlinear programming; augmented Lagrangian; numerical experiments;
D O I
10.1007/s10107-004-0564-1
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
The low-rank semidefinite programming problem LRSDPr is a restriction of the semidefinite programming problem SDP in which a bound r is imposed on the rank of X, and it is well known that LRSDPr is equivalent to SDP if r is not too small. In this paper, we classify the local minima of LRSDPr and prove the optimal convergence of a slight variant of the successful, yet experimental, algorithm of Burer and Monteiro [5], which handles LRSDPr via the nonconvex change of variables X = RRT. In addition, for particular problem classes, we describe a practical technique for obtaining lower bounds on the optimal solution value during the execution of the algorithm. Computational results are presented on a set of combinatorial optimization relaxations, including some of the largest quadratic assignment SDPs solved to date.
引用
收藏
页码:427 / 444
页数:18
相关论文
共 27 条
[1]   Primal-dual interior-point methods for semidefinite programming: Convergence rates, stability and numerical results [J].
Alizadeh, F ;
Haeberly, JPA ;
Overton, ML .
SIAM JOURNAL ON OPTIMIZATION, 1998, 8 (03) :746-768
[2]   Recent advances in the solution of quadratic assignment problems [J].
Anstreicher, KM .
MATHEMATICAL PROGRAMMING, 2003, 97 (1-2) :27-42
[3]   CONES OF DIAGONALLY DOMINANT MATRICES [J].
BARKER, GP ;
CARLSON, D .
PACIFIC JOURNAL OF MATHEMATICS, 1975, 57 (01) :15-32
[5]   A nonlinear programming algorithm for solving semidefinite programs via low-rank factorization [J].
Burer, S ;
Monteiro, RDC .
MATHEMATICAL PROGRAMMING, 2003, 95 (02) :329-357
[6]   Solving a class of semidefinite programs via nonlinear programming [J].
Burer, S ;
Monteiro, RDC ;
Zhang, Y .
MATHEMATICAL PROGRAMMING, 2002, 93 (01) :97-122
[7]   QAPLIB-A QUADRATIC ASSIGNMENT PROBLEM LIBRARY [J].
BURKARD, RE ;
KARISCH, S ;
RENDL, F .
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 1991, 55 (01) :115-119
[8]  
De Simone C., 1989, DISCRETE MATH, V79, P71
[9]   A spectral bundle method for semidefinite programming [J].
Helmberg, C ;
Rendl, F .
SIAM JOURNAL ON OPTIMIZATION, 2000, 10 (03) :673-696
[10]   An interior-point method for semidefinite programming [J].
Helmberg, C ;
Rendl, F ;
Vanderbei, RJ ;
Wolkowicz, H .
SIAM JOURNAL ON OPTIMIZATION, 1996, 6 (02) :342-361