Reproduce “Efficient Rank Minimization to Tighten Semidefinite Programming for Unconstrained Binary Quadratic Optimization” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Roman Pogodin, Mikhail Krechetov, Yury Maximov
The authors develop a rank-reduction method for semidefinite programming solutions to unconstrained binary quadratic optimization, aiming to improve randomized rounding. It minimizes smooth rank surrogates with updates designed to preserve positive semidefiniteness without cone projection under problem-specific constraints. On Gset graph #12, rank fell from 9 to 8 and the rounded cut rose from 532 to 534; across tested instances, cuts improved on approximately half.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.