Sparse Semidefinite Programs with Guaranteed NearLinear Time Complexity via Dualized Clique Tree Conversion
Abstract
Clique tree conversion solves largescale semidefinite programs by splitting an $n\times n$ matrix variable into up to $n$ smaller matrix variables, each representing a principal submatrix. Its fundamental weakness is the need to introduce \emph{overlap constraints} that enforce agreement between different matrix variables, because these can result in dense coupling. In this paper, we show that by dualizing the clique tree conversion, the coupling due to the overlap constraints is guaranteed to be sparse over dense blocks, with a block sparsity pattern that coincides with the adjacency matrix of a tree. In two classes of semidefinite programs with favorable sparsity patterns that encompass the MAXCUT and MAX $k$CUT relaxations, the Lovasz Theta problem, and the AC optimal power flow relaxation, we prove that the periteration cost of an interiorpoint method is linear $O(n)$ time and memory, so an $\epsilon$accurate and $\epsilon$feasible iterate is obtained after $O(\sqrt{n}\log(1/\epsilon))$ iterations in nearlinear $O(n^{1.5}\log(1/\epsilon))$ time. We confirm our theoretical insights with numerical results on semidefinite programs as large as $n=13659$.
 Publication:

arXiv eprints
 Pub Date:
 October 2017
 arXiv:
 arXiv:1710.03475
 Bibcode:
 2017arXiv171003475Z
 Keywords:

 Mathematics  Optimization and Control
 EPrint:
 [v1] appeared in IEEE CDC 2018