Strong formulations for quadratic optimization with M-matrices and indicator variables

We study quadratic optimization with indicator variables and an M-matrix, i.e., a PSD matrix with non-positive off-diagonal entries, which arises directly in image segmentation and portfolio optimization with transaction costs, as well as a substructure of general quadratic optimization problems. We prove, under mild assumptions, that the minimization problem is solvable in polynomial time by showing its equivalence to a submodular minimization problem. To strengthen the formulation, we decompose the quadratic function into a sum of simple quadratic functions with at most two indicator variables each, and provide the convex-hull descriptions of these sets. We also describe strong conic quadratic valid inequalities. Preliminary computational experiments indicate that the proposed inequalities can substantially improve the strength of the continuous relaxations with respect to the standard perspective reformulation.

[1]  Juan Pablo Vielma,et al.  Intersection cuts for nonlinear integer programming: convexification techniques for structured sets , 2013, Mathematical Programming.

[2]  Sven Leyffer,et al.  Minotaur: a mixed-integer nonlinear optimization toolkit , 2021, Math. Program. Comput..

[3]  Henry Wolkowicz,et al.  Convex Relaxations of (0, 1)-Quadratic Programming , 1995, Math. Oper. Res..

[4]  Sinan Gürel,et al.  A strong conic quadratic reformulation for machine-job assignment with controllable processing times , 2009, Oper. Res. Lett..

[5]  Dorit S. Hochbaum,et al.  A Cut-Based Algorithm for the Nonlinear Dual of the Minimum Cost Network Flow Problem , 2004, Algorithmica.

[6]  Xiaojin Zheng,et al.  Quadratic Convex Reformulations for Semicontinuous Quadratic Programming , 2017, SIAM J. Optim..

[7]  J. Keilson,et al.  Markov chains and M-matrices: Inequalities and equalities , 1973 .

[8]  E. Balas Disjunctive programming and a hierarchy of relaxations for discrete optimization problems , 1985 .

[9]  Jeff T. Linderoth,et al.  On Valid Inequalities for Quadratic Programming with Continuous Variables and Binary Indicators , 2013, IPCO.

[10]  Gérard Cornuéjols,et al.  Optimization Methods in Finance: Stochastic programming: theory and algorithms , 2006 .

[11]  Charles K. Sestok,et al.  Sparse Filter Design Under a Quadratic Constraint: Low-Complexity Algorithms , 2013, IEEE Transactions on Signal Processing.

[12]  James R. Luedtke,et al.  Strong Convex Nonlinear Relaxations of the Pooling Problem , 2018, SIAM J. Optim..

[13]  Sanjay Mehrotra,et al.  A branch-and-cut method for 0-1 mixed convex programming , 1999, Math. Program..

[14]  Hassan L. Hijazi,et al.  Mixed-integer nonlinear programs featuring “on/off” constraints , 2012, Comput. Optim. Appl..

[15]  James R. Luedtke,et al.  Some results on the strength of relaxations of multilinear functions , 2012, Math. Program..

[16]  Alper Atamtürk,et al.  Submodularity in Conic Quadratic Mixed 0-1 Optimization , 2017, Oper. Res..

[17]  Claudio Gentile,et al.  Perspective cuts for a class of convex 0–1 mixed integer programs , 2006, Math. Program..

[18]  László Lovász,et al.  Submodular functions and convexity , 1982, ISMP.

[19]  Javier F. Pena,et al.  Optimization Methods in Finance: Stochastic programming: theory and algorithms , 2006 .

[20]  H. D. Ratliff,et al.  Minimum cuts and related problems , 1975, Networks.

[21]  Marco Molinaro,et al.  Bounding the gap between the McCormick relaxation and the convex hull for bilinear functions , 2015, Math. Program..

[22]  J. Hiriart-Urruty,et al.  Convex analysis and minimization algorithms , 1993 .

[23]  Alper Atamtürk,et al.  Lifted polymatroid inequalities for mean-risk optimization with indicator variables , 2019, J. Glob. Optim..

[24]  D. Hochbaum Multi-Label Markov Random Fields as an Efficient and Effective Tool for Image Segmentation, Total Variations and Regularization , 2013 .

[25]  R. Plemmons M-matrix characterizations.I—nonsingular M-matrices , 1977 .

[26]  D. Bertsimas,et al.  Best Subset Selection via a Modern Optimization Lens , 2015, 1507.03133.

[27]  Jack Edmonds,et al.  Submodular Functions, Matroids, and Certain Polyhedra , 2001, Combinatorial Optimization.

[28]  Julio C. Góez,et al.  A Conic Representation of the Convex Hull of Disjunctive Sets and Conic Cuts for Integer Second Order Cone Optimization , 2015 .

[29]  Daniel Bienstock,et al.  Computational study of a family of mixed-integer quadratic programming problems , 1995, Math. Program..

[30]  Pierre Bonami,et al.  On mathematical programming with indicator constraints , 2015, Math. Program..

[31]  Sercan Yildiz,et al.  Two-term disjunctions on the second-order cone , 2014, IPCO.

[32]  Duan Li,et al.  Cardinality Constrained Linear-Quadratic Optimal Control , 2011, IEEE Transactions on Automatic Control.

[33]  Juan Pablo Vielma Small and strong formulations for unions of convex sets from the Cayley embedding , 2017, Mathematical Programming.

[34]  Nicholas Young The rate of convergence of a matrix power series , 1981 .

[35]  Daniel Bienstock,et al.  Cutting-Planes for Optimization of Convex Functions over Nonconvex Sets , 2014, SIAM J. Optim..

[36]  Alper Atamtürk,et al.  Cuts for Conic Mixed-Integer Programming , 2007, IPCO.

[37]  Alper Atamtürk,et al.  Network design with probabilistic capacities , 2017, Networks.

[38]  Petru L. Ivănescu,et al.  Pseudo-Boolean programming , 1965 .

[39]  Oktay Günlük,et al.  Perspective reformulations of mixed integer nonlinear programs with indicator variables , 2010, Math. Program..

[40]  Franklin T. Luk,et al.  Quadratic Programming with M-Matrices , 1979 .

[41]  Stephen P. Boyd,et al.  Portfolio optimization with linear and fixed transaction costs , 2007, Ann. Oper. Res..

[42]  M. L. Fisher,et al.  An analysis of approximations for maximizing submodular set functions—I , 1978, Math. Program..

[43]  Kurt M. Anstreicher,et al.  On convex relaxations for quadratically constrained quadratic programming , 2012, Math. Program..

[44]  Jeff T. Linderoth,et al.  Quadratic cone cutting surfaces for quadratic programs with on-off constraints , 2017, Discret. Optim..

[45]  James B. Orlin,et al.  A faster strongly polynomial time algorithm for submodular function minimization , 2007, Math. Program..

[46]  Claudio Gentile,et al.  Decompositions of Semidefinite Matrices and the Perspective Reformulation of Nonseparable Quadratic Programs , 2020, Math. Oper. Res..

[47]  Olga Veksler,et al.  Fast Approximate Energy Minimization via Graph Cuts , 2001, IEEE Trans. Pattern Anal. Mach. Intell..

[48]  Sebastián Ceria,et al.  Convex programming for disjunctive convex optimization , 1999, Math. Program..