REVIEW 5 cited by
A low-rank augmented Lagrangian method for large-scale semidefinite programming based on a hybrid convex-nonconvex approach
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
abstract
This paper introduces HALLaR, a new first-order method for solving large-scale semidefinite programs (SDPs) with bounded domain. HALLaR is an inexact augmented Lagrangian (AL) method where the AL subproblems are solved by a novel hybrid low-rank (HLR) method. The recipe behind HLR is based on two key ingredients: 1) an adaptive inexact proximal point method with inner acceleration; 2) Frank-Wolfe steps to escape from spurious local stationary points. In contrast to the low-rank method of Burer and Monteiro, HALLaR finds a near-optimal solution (with provable complexity bounds) of SDP instances satisfying strong duality. Computational results comparing HALLaR to state-of-the-art solvers on several large SDP instances arising from maximum stable set, phase retrieval, and matrix completion show that the former finds higher accurate solutions in substantially less CPU time than the latter ones. For example, in less than 20 minutes, HALLaR can solve a maximum stable set SDP instance with dimension pair $(n,m)\approx (10^6,10^7)$ within $10^{-5}$ relative precision.
Forward citations
Cited by 5 Pith papers
-
New Understandings and Computation on Augmented Lagrangian Methods for Low-Rank Semidefinite Programming
Augmented Lagrangian subproblems inherit low-rankness, strict complementarity, and quadratic growth from a primal simple SDP, making Burer-Monteiro gradient descent converge linearly.
-
Solving Imperfect-Recall Games via Sum-of-Squares Optimization
Moment-SOS hierarchies provably compute behavioral optima/equilibria in imperfect-recall games, with exact convergence at level ℓ+1 for non-absentminded single-player games.
-
RiNNAL+: a Riemannian ALM Solver for SDP-RLT Relaxations of Mixed-Binary Quadratic Programs
DNN and SDP-RLT relaxations of mixed-binary quadratic programs are proved to give identical bounds, and the new hybrid Riemannian solver RiNNAL+ solves the smaller SDP-RLT form at n = 5000, typically 10 to 100 times f...
-
A Bundle-based Augmented Lagrangian Framework: Algorithm, Convergence, and Primal-dual Principles
BALA, a single-loop bundle-based augmented Lagrangian algorithm, achieves sublinear convergence for convex constrained problems and linear convergence for a class of conic programs including semidefinite programs.
-
Maximal entropy in the moment body
After preconditioning the defining linear map, global minimization of the dual log-partition function certifies moment body membership, and L-BFGS handles dense n=m=1000 instances in seconds.
Discussion (0). Continue with ORCID to comment.