Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:1603.00347 · INFORMS journal on computing

Using a Conic Bundle Method to Accelerate Both Phases of a Quadratic Convex Reformulation

Alain Billionnet, Sourour Elloumi, Amélie Lambert, Angelika Wiegele

Published Apr 12, 2017• 17 citations
arXivHTMLPDF

Abstract

We present algorithm MIQCR-CB that is an advancement of MIQCR. MIQCR is a method for solving mixed-integer quadratic programs and works in two phases: the first phase determines an equivalent quadratic formulation with a convex objective function by solving a semidefinite problem (SDP); in the second phase, the equivalent formulation is solved by a standard solver. Because the reformulation relies on the solution of a large-scale semidefinite program, it is not tractable by existing semidefinite solvers even for medium-sized problems. To surmount this difficulty, we present in MIQCR-CB a subgradient algorithm within a Lagrangian duality framework for solving (SDP) that substantially speeds up the first phase. Moreover, this algorithm leads to a reformulated problem of smaller size than the one obtained by the original MIQCR method, which results in a shorter time for solving the second phase. We present extensive computational results to show the efficiency of our algorithm. First, we apply MIQCR-CB to the k-cluster problem that can be formulated by a binary quadratic program. As an illustration of the efficiency of our new algorithm, for instances of size 80 and of density 25%, MIQCR-CB is on average 78 times faster for phase 1 and 24 times faster for phase 2 than the original MIQCR. We also compare MIQCR-CB with QCR and with BiqCrunch, two methods devoted to binary quadratic programming. We show that MIQCR-CB is able to solve most of the 225 considered instances within three hours of CPU time. We also present experiments on two classes of general integer instances where we compare MIQCR-CB with MIQCR, Couenne, and Cplex12.6. We demonstrate the significant improvement over the original MIQCR approach. Finally, we show that MIQCR-CB is able to solve almost all of the considered instances, whereas Couenne and Cplex12.6 are not able to solve half of them.

DOI
10.1287/ijoc.2016.0731

Related papers

Exact Convex Reformulations of Binary Quadratic Programs from Facially Reduced Semidefinite Relaxations

Hao Hu, Mingming Xu

We develop MIQCR-FR, a framework for exact convex reformulation from facially reduced semidefinite relaxations of equality-constrained binary quadratic problems. Standard mixed-integer quadratic convex reformulation (MIQCR) uses dual multipliers of the original relaxation. Facial reduction (FR) reduces the semidefinite matrix order, but recovery of the original dual multipliers is not guaranteed in general. For this class of problems, we identify how retaining the linear equalities and their products with each binary variable enables this recovery. An established dual recovery formula gives an explicit extension of every feasible solution of the reduced dual to the original dual, preserving its bound. For approximate multiplier estimates, we give a spectral shift restoring dual feasibility and two constructions optimizing different bounds within a restricted multiplier family. Maximizing the dual bound yields a trust-region subproblem; maximizing the continuous MIQCR bound requires only a reduced eigenvalue calculation. The latter construction enforces convexity directly, rather than full dual feasibility. We implement it using multiplier estimates from the alternating direction method of multipliers (ADMM). On 431 instances from six problem families, the complete method certifies 296 optima, compared with 156 for the conic-bundle method MIQCR-CB and 187 for direct Gurobi under a common time budget.

arXiv:2610.11154•Oct 8, 2026•0 citations

Global solution of non-convex quadratically constrained quadratic programs

Sourour Elloumi, Amélie Lambert

The class of mixed-integer quadratically constrained quadratic programs (QCQP) consists of minimizing a quadratic function under quadratic constraints where the variables could be integer or continuous. On a previous paper we introduced a method called MIQCR for solving QCQPs with the following restriction: all quadratic sub-functions of purely continuous variables are already convex. In this paper, we propose an extension of MIQCR which applies to any QCQP. Let (P) be a QCQP. Our approach to solve (P) is first to build an equivalent mixed-integer quadratic problem (P∗). This equivalent problem (P∗) has a quadratic convex objective function, linear constraints, and additional variables y that are meant to satisfy the additional quadratic constraints y=xxT, where x are the initial variables of problem (P). We then propose to solve (P∗) by a branch-and-bound algorithm based on the relaxation of the additional quadratic constraints and of the integrality constraints. This type of branching is known as spatial branch-and-bound. Computational experiences are carried out on a total of 325 instances. The results show that the solution time of most of the considered instances is improved by our method in comparison with the recent implementation of QuadProgBB, and with the solvers Cplex, Couenne, Scip, BARON and GloMIQO.

Optimization methods & software•Jul 28, 2017•25 citations

Recent advances in solving some optimization problems in graphs by quadratic programming

Sourour Elloumi, Amélie Lambert

We review Quadratic Convex Reformulation (QCR) for quadratic pro-grams with general integer variables. This solution 2-phase approachconsist in first reformulating the quadratic program into an equivalentother problem having a convex ob jective function. The second phaserelies on MIP solvers that solve the reformulated problem by standardbranch-and-b ound. Then, we consider some graph partitioning problemsthat can be formulated as quadratic programs with binary variables. Weshow many enhancements of the standard QCR method that efficientlysolve graph partitioning problems

•Jul 7, 2014•0 citations

Improving the Performance of MIQP Solvers for Quadratic Programs with Cardinality and Minimum Threshold Constraints: A Semidefinite Program Approach

Xiaojin Zheng, Xiaoling Sun, Duan Li

We consider in this paper quadratic programming problems with cardinality and minimum threshold constraints that arise naturally in various real-world applications such as portfolio selection and subset selection in regression. This class of problems can be formulated as mixed-integer 0-1 quadratic programs. We propose a new semidefinite program (SDP) approach for computing the “best” diagonal decomposition that gives the tightest continuous relaxation of the perspective reformulation of the problem. We also give an alternative way of deriving the perspective reformulation by applying a special Lagrangian decomposition scheme to the diagonal decomposition of the problem. This derivation can be viewed as a “dual” method to the convexification method employing the perspective function on semicontinuous variables. Computational results show that the proposed SDP approach can be advantageous for improving the performance of mixed-integer quadratic programming solvers when applied to the perspective reformulations of the problem.

INFORMS journal on computing•May 19, 2014•69 citations

Quadratic 0–1 programming: Tightening linear or quadratic convex reformulation by use of relaxations

Alain Billionnet, Sourour Elloumi, Marie-Christine Plateau

Many combinatorial optimization problems can be formulated as the minimization of a 0–1 quadratic function subject to linear constraints. In this paper, we are interested in the exact solution of this problem through a two-phase general scheme. The first phase consists in reformulating the initial problem either into a compact mixed integer linear program or into a 0–1 quadratic convex program. The second phase simply consists in submitting the reformulated problem to a standard solver. The efficiency of this scheme strongly depends on the quality of the reformulation obtained in phase 1. We show that a good compact linear reformulation can be obtained by solving a continuous linear relaxation of the initial problem. We also show that a good quadratic convex reformulation can be obtained by solving a semidefinite relaxation. In both cases, the obtained reformulation profits from the quality of the underlying relaxation. Hence, the proposed scheme gets around, in a sense, the difficulty to incorporate these costly relaxations in a branch-and-bound algorithm.

RAIRO - Operations Research•Apr 1, 2008•22 citations

Large-Scale Binary Quadratic Optimization Using Semidefinite Relaxation and Applications

Peng Wang, Chunhua Shen, Anton van den Hengel +1 more

In computer vision, many problems can be formulated as binary quadratic programs (BQPs), which are in general NP hard. Finding a solution when the problem is of large size to be of practical interest typically requires relaxation. Semidefinite relaxation usually yields tight bounds, but its computational complexity is high. In this work, we present a semidefinite programming (SDP) formulation for BQPs, with two desirable properties. First, it produces similar bounds to the standard SDP formulation. Second, compared with the conventional SDP formulation, the proposed SDP formulation leads to a considerably more efficient and scalable dual optimization approach. We then propose two solvers, namely, quasi-Newton and smoothing Newton methods, for the simplified dual problem. Both of them are significantly more efficient than standard interior-point methods. Empirically the smoothing Newton solver is faster than the quasi-Newton solver for dense or medium-sized problems, while the quasi-Newton solver is preferable for large sparse/structured problems.

arXiv:1411.7564•IEEE Transactions on Pattern Analysis and Machine Intelligence•Mar 11, 2016•59 citations
We present algorithm MIQCR-CB that is an advancement of MIQCR. MIQCR is a method for solving mixed-integer quadratic programs and works in two phases: the first phase determines an equivalent quadratic formulation with a convex objective function by solving a semidefinite problem (SDP); in the second phase, the equivalent formulation is solved by a standard solver. Because the reformulation relies on the solution of a large-scale semidefinite program, it is not tractable by existing semidefinite solvers even for medium-sized problems. To surmount this difficulty, we present in MIQCR-CB a subgradient algorithm within a Lagrangian duality framework for solving (SDP) that substantially speeds up the first phase. Moreover, this algorithm leads to a reformulated problem of smaller size than the one obtained by the original MIQCR method, which results in a shorter time for solving the second phase. We present extensive computational results to show the efficiency of our algorithm. First, we apply MIQCR-CB to the k-cluster problem that can be formulated by a binary quadratic program. As an illustration of the efficiency of our new algorithm, for instances of size 80 and of density 25%, MIQCR-CB is on average 78 times faster for phase 1 and 24 times faster for phase 2 than the original MIQCR. We also compare MIQCR-CB with QCR and with BiqCrunch, two methods devoted to binary quadratic programming. We show that MIQCR-CB is able to solve most of the 225 considered instances within three hours of CPU time. We also present experiments on two classes of general integer instances where we compare MIQCR-CB with MIQCR, Couenne, and Cplex12.6. We demonstrate the significant improvement over the original MIQCR approach. Finally, we show that MIQCR-CB is able to solve almost all of the considered instances, whereas Couenne and Cplex12.6 are not able to solve half of them.
We develop MIQCR-FR, a framework for exact convex reformulation from facially reduced semidefinite relaxations of equality-constrained binary quadratic problems. Standard mixed-integer quadratic convex reformulation (MIQCR) uses dual multipliers of the original relaxation. Facial reduction (FR) reduces the semidefinite matrix order, but recovery of the original dual multipliers is not guaranteed in general. For this class of problems, we identify how retaining the linear equalities and their products with each binary variable enables this recovery. An established dual recovery formula gives an explicit extension of every feasible solution of the reduced dual to the original dual, preserving its bound. For approximate multiplier estimates, we give a spectral shift restoring dual feasibility and two constructions optimizing different bounds within a restricted multiplier family. Maximizing the dual bound yields a trust-region subproblem; maximizing the continuous MIQCR bound requires only a reduced eigenvalue calculation. The latter construction enforces convexity directly, rather than full dual feasibility. We implement it using multiplier estimates from the alternating direction method of multipliers (ADMM). On 431 instances from six problem families, the complete method certifies 296 optima, compared with 156 for the conic-bundle method MIQCR-CB and 187 for direct Gurobi under a common time budget.
The class of mixed-integer quadratically constrained quadratic programs (QCQP) consists of minimizing a quadratic function under quadratic constraints where the variables could be integer or continuous. On a previous paper we introduced a method called MIQCR for solving QCQPs with the following restriction: all quadratic sub-functions of purely continuous variables are already convex. In this paper, we propose an extension of MIQCR which applies to any QCQP. Let (P) be a QCQP. Our approach to solve (P) is first to build an equivalent mixed-integer quadratic problem (P∗). This equivalent problem (P∗) has a quadratic convex objective function, linear constraints, and additional variables y that are meant to satisfy the additional quadratic constraints y=xxT, where x are the initial variables of problem (P). We then propose to solve (P∗) by a branch-and-bound algorithm based on the relaxation of the additional quadratic constraints and of the integrality constraints. This type of branching is known as spatial branch-and-bound. Computational experiences are carried out on a total of 325 instances. The results show that the solution time of most of the considered instances is improved by our method in comparison with the recent implementation of QuadProgBB, and with the solvers Cplex, Couenne, Scip, BARON and GloMIQO.
We review Quadratic Convex Reformulation (QCR) for quadratic pro-grams with general integer variables. This solution 2-phase approachconsist in first reformulating the quadratic program into an equivalentother problem having a convex ob jective function. The second phaserelies on MIP solvers that solve the reformulated problem by standardbranch-and-b ound. Then, we consider some graph partitioning problemsthat can be formulated as quadratic programs with binary variables. Weshow many enhancements of the standard QCR method that efficientlysolve graph partitioning problems
We consider in this paper quadratic programming problems with cardinality and minimum threshold constraints that arise naturally in various real-world applications such as portfolio selection and subset selection in regression. This class of problems can be formulated as mixed-integer 0-1 quadratic programs. We propose a new semidefinite program (SDP) approach for computing the “best” diagonal decomposition that gives the tightest continuous relaxation of the perspective reformulation of the problem. We also give an alternative way of deriving the perspective reformulation by applying a special Lagrangian decomposition scheme to the diagonal decomposition of the problem. This derivation can be viewed as a “dual” method to the convexification method employing the perspective function on semicontinuous variables. Computational results show that the proposed SDP approach can be advantageous for improving the performance of mixed-integer quadratic programming solvers when applied to the perspective reformulations of the problem.
Many combinatorial optimization problems can be formulated as the minimization of a 0–1 quadratic function subject to linear constraints. In this paper, we are interested in the exact solution of this problem through a two-phase general scheme. The first phase consists in reformulating the initial problem either into a compact mixed integer linear program or into a 0–1 quadratic convex program. The second phase simply consists in submitting the reformulated problem to a standard solver. The efficiency of this scheme strongly depends on the quality of the reformulation obtained in phase 1. We show that a good compact linear reformulation can be obtained by solving a continuous linear relaxation of the initial problem. We also show that a good quadratic convex reformulation can be obtained by solving a semidefinite relaxation. In both cases, the obtained reformulation profits from the quality of the underlying relaxation. Hence, the proposed scheme gets around, in a sense, the difficulty to incorporate these costly relaxations in a branch-and-bound algorithm.
In computer vision, many problems can be formulated as binary quadratic programs (BQPs), which are in general NP hard. Finding a solution when the problem is of large size to be of practical interest typically requires relaxation. Semidefinite relaxation usually yields tight bounds, but its computational complexity is high. In this work, we present a semidefinite programming (SDP) formulation for BQPs, with two desirable properties. First, it produces similar bounds to the standard SDP formulation. Second, compared with the conventional SDP formulation, the proposed SDP formulation leads to a considerably more efficient and scalable dual optimization approach. We then propose two solvers, namely, quasi-Newton and smoothing Newton methods, for the simplified dual problem. Both of them are significantly more efficient than standard interior-point methods. Empirically the smoothing Newton solver is faster than the quasi-Newton solver for dense or medium-sized problems, while the quasi-Newton solver is preferable for large sparse/structured problems.