Computable representations for convex hulls of low-dimensional quadratic forms

被引:0
|
作者
Kurt M. Anstreicher
Samuel Burer
机构
[1] University of Iowa,Department of Management Sciences
来源
Mathematical Programming | 2010年 / 124卷
关键词
Quadratic form; Convex hull; Convex envelope; Global optimization; Semidefinite programming; 90C20; 90C22; 90C26;
D O I
暂无
中图分类号
学科分类号
摘要
Let \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathcal{C}}$$\end{document} be the convex hull of points \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{\{{1 \choose x}{1 \choose x}^T \,|\, x\in \mathcal{F}\subset \Re^n\}}}$$\end{document}. Representing or approximating \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathcal{C}}$$\end{document} is a fundamental problem for global optimization algorithms based on convex relaxations of products of variables. We show that if n ≤ 4 and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathcal{F}}$$\end{document} is a simplex, then \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathcal{C}}$$\end{document} has a computable representation in terms of matrices X that are doubly nonnegative (positive semidefinite and componentwise nonnegative). We also prove that if n = 2 and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathcal{F}}$$\end{document} is a box, then \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathcal{C}}$$\end{document} has a representation that combines semidefiniteness with constraints on product terms obtained from the reformulation-linearization technique (RLT). The simplex result generalizes known representations for the convex hull of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{\{(x_1, x_2, x_1x_2)\,|\, x\in\mathcal{F}\}}}$$\end{document} when \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathcal{F}\subset\Re^2}$$\end{document} is a triangle, while the result for box constraints generalizes the well-known fact that in this case the RLT constraints generate the convex hull of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{\{(x_1, x_2, x_1x_2)\,|\, x\in\mathcal{F}\}}}$$\end{document}. When n = 3 and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathcal{F}}$$\end{document} is a box, we show that a representation for \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathcal{C}}$$\end{document} can be obtained by utilizing the simplex result for n = 4 in conjunction with a triangulation of the 3-cube.
引用
收藏
页码:33 / 43
页数:10
相关论文
共 50 条
  • [31] Representations of indefinite quadratic forms
    Hsia, JS
    Shao, YY
    Xu, F
    JOURNAL FUR DIE REINE UND ANGEWANDTE MATHEMATIK, 1998, 494 : 129 - 140
  • [32] Polar representations of compact groups and convex hulls of their orbits
    Gichev, V. M.
    DIFFERENTIAL GEOMETRY AND ITS APPLICATIONS, 2010, 28 (05) : 608 - 614
  • [33] MINIMIZING A LOW-DIMENSIONAL CONVEX FUNCTION OVER A HIGH-DIMENSIONAL CUBE
    Hunkenschroder, Christoph
    Pokutta, Sebastian
    Weismantel, Robert
    SIAM JOURNAL ON OPTIMIZATION, 2023, 33 (02) : 538 - 552
  • [34] Low-dimensional representations of the three component loop braid group
    Bruillard, Paul
    Chang, Liang
    Hong, Seung-Moon
    Plavnik, Julia Yael
    Rowell, Eric C.
    Sun, Michael Yuan
    JOURNAL OF MATHEMATICAL PHYSICS, 2015, 56 (11)
  • [35] Low-Dimensional Text Representations for Sentiment Analysis NLP Tasks
    Akritidis L.
    Bozanis P.
    SN Computer Science, 4 (5)
  • [36] Low-dimensional linear representations of Aut Fn, n ≥ 3
    Potapchik, A
    Rapinchuk, A
    TRANSACTIONS OF THE AMERICAN MATHEMATICAL SOCIETY, 2000, 352 (03) : 1437 - 1451
  • [37] APPROXIMATE REPRESENTATIONS, APPROXIMATE HOMOMORPHISMS, AND LOW-DIMENSIONAL EMBEDDINGS OF GROUPS
    Moore, Cristopher
    Russell, Alexander
    SIAM JOURNAL ON DISCRETE MATHEMATICS, 2015, 29 (01) : 182 - 197
  • [38] Minimal linear representations of the low-dimensional nilpotent Lie algebras
    Benjumea, J. C.
    Nunez, J.
    Tenorio, A. F.
    MATHEMATICA SCANDINAVICA, 2008, 102 (01) : 17 - 26
  • [39] Low-dimensional representations of shaded surfaces under varying illumination
    Nillius, P
    Eklundh, JO
    2003 IEEE COMPUTER SOCIETY CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, VOL II, PROCEEDINGS, 2003, : 185 - 192
  • [40] Generating Low-dimensional, Nonlinear Process Representations by Ordered Features
    Fischer, Susanne
    Hensgen, Onno
    Elshaabiny, Moustafa
    Link, Norbert
    IFAC PAPERSONLINE, 2015, 48 (03): : 1037 - 1042