Abstract:We consider a quadratic program with a few negative eigenvalues (QP-r-NE) subject to linear and convex quadratic constraints that covers many applications and is known to be NP-hard even with one negative eigenvalue (QP1NE). In this paper, we first introduce a new global algorithm (ADMBB), which integrates several simple optimization techniques such as alternative direction method, and branch-and-bound, to find a globally optimal solution to the underlying QP within a pre-specified ϵ\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$$\epsilon $$\end{document}-tolerance. We establish the convergence of the ADMBB algorithm and estimate its complexity. Second, we develop a global search algorithm (GSA) for QP1NE that can locate an optimal solution to QP1NE within ϵ\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$$\epsilon $$\end{document}-tolerance and estimate the worst-case complexity bound of the GSA. Preliminary numerical results demonstrate that the ADMBB algorithm can effectively find a global optimal solution to large-scale QP-r-NE instances when r≤10\documentclass[12pt]{minimal}\usepackage{amsmath}\usepackage{wasysym}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{amsbsy}\usepackage{mathrsfs}\usepackage{upgreek}\setlength{\oddsidemargin}{-69pt}\begin{document}$$r\le 10$$\end{document}, and the GSA outperforms the ADMBB for most of the tested QP1NE instances. The software reviewed as part of this submission was given the DOI (digital object identifier) 10.5281/zenodo.1344739.

A Certifiably Globally Optimal Solution to Generalized Essential Matrix Estimation

Toward Globally Optimal State Estimation Using Automatically Tightened Semidefinite Relaxations

Accelerating Globally Optimal Consensus Maximization in Geometric Vision

Globally Optimal Relative Pose Estimation for Multi-Camera Systems with Known Gravity Direction

An efficient nonconvex reformulation of stagewise convex optimization problems

Convex relaxation for the generalized maximum-entropy sampling problem

Distributed Certifiably Correct Pose-Graph Optimization

General Constrained Matrix Optimization

Distributed Nonconvex Optimization: Gradient-free Iterations and $ε$-Globally Optimal Solution

Quadratic Pose Estimation Problems: Globally Optimal Solutions, Solvability/Observability Analysis, and Uncertainty Description

Certifiably Optimal Rotation and Pose Estimation Based on the Cayley Map

Spectrally Constrained Optimization

Do not Omit Local Minimizer: a Complete Solution for Pose Estimation from 3D Correspondences

Distributed Nonconvex Optimization: Gradient-free Iterations and $\Epsilon$-Globally Optimal Solution

FracGM: A Fast Fractional Programming Technique for Geman-McClure Robust Estimator

One Ring to Rule Them All: Certifiably Robust Geometric Perception with Outliers

A Whole New Ball Game: A Primal Accelerated Method for Matrix Games and Minimizing the Maximum of Smooth Functions

An efficient active-set method with applications to sparse approximations and risk minimization

New global algorithms for quadratic programming with a few negative eigenvalues based on alternative direction method and convex relaxation