## Abstract

We present effective linear programming based computational techniques for solving nonconvex quadratic programs with box constraints (BoxQP). We first observe that known cutting planes obtained from the Boolean Quadric Polytope (BQP) are computationally effective at reducing the optimality gap of BoxQP. We next show that the Chvátal–Gomory closure of the BQP is given by the odd-cycle inequalities even when the underlying graph is not complete. By using these cutting planes in a spatial branch-and-cut framework, together with a common integrality-based preprocessing technique and a particular convex quadratic relaxation, we develop a solver that can effectively solve a well-known family of test instances. Our linear programming based solver is competitive with SDP-based state of the art solvers on small instances and sparse instances. Most of our computational techniques have been implemented in the recent version of CPLEX and have led to significant performance improvements on nonconvex quadratic programs with linear constraints.

This is a preview of subscription content, access via your institution.

## References

- 1.
An, L.T.H., Tao, P.D.: A branch and bound method via d.c. optimization algorithms and ellipsoidal technique for box constrained nonconvex quadratic problems. J. Glob. Optim.

**13**, 171–206 (1998) - 2.
Andersen, M., Dahl, J., Vandenberghe, L.: CVXOPT user’s guide, release 1.1.8 (2015)

- 3.
Anstreicher, K.: On convex relaxations for quadratically constrained quadratic programming. Math. Program.

**136**, 233–251 (2012) - 4.
Anstreicher, K., Burer, S.: Computable representations for convex hulls of low-dimensional quadratic forms. Math. Program.

**124**, 33–43 (2010) - 5.
Anstreicher, K.M.: Semidefinite programming versus the reformulation-linearization technique for nonconvex quadratically constrained quadratic programming. J. Glob. Optim.

**43**(2), 471–484 (2008) - 6.
Barahona, F.: On cuts and matchings in planar graphs. Math. Program.

**60**, 53,58 (1993) - 7.
Barahona, F., Grötschel, M., Jünger, M., Reinelt, G.: Experiments in quadratic 01 programming. Math. Program.

**44**, 127–137 (1989) - 8.
Barahona, F., Mahjoub, A.: On the cut polytope. Math. Program.

**36**, 157–173 (1986) - 9.
Bliek, C., Bonami, P., Lodi, A.: Solving mixed-integer quadratic programming problems with IBM-CPLEX: a progress report. In: Proceedings of the Twenty-Sixth RAMP Symposium, pp. 171–180 (2014)

- 10.
Boros, E., Crama, Y., Hammer, P.L.: Chvátal cuts and odd cycle inequalities in quadratic 0–1 optimization. SIAM J. Discrete Math.

**5**(2), 163–177 (1992) - 11.
Boros, E., Hammer, P.L.: Cut-polytopes, Boolean quadric polytopes and nonnegative quadratic pseudo-Boolean functions. Math. Oper. Res.

**18**(1), 245–253 (1993) - 12.
Burer, S., Monteiro, D.R.: A nonlinear programming algorithm for solving semidefinite programs via low-rank factorization. Math. Program.

**95**(2), 329–357 (2003). https://doi.org/10.1007/s10107-002-0352-8 - 13.
Burer, S.: Optimizing a polyhedral-semidefinite relaxation of completely positive programs. Math. Program. Comput.

**2**(1), 119 (2010) - 14.
Burer, S., Chen, J.: Globally solving nonconvex quadratic programming problems via completely positive programming. Math. Program. Comput.

**4**(1), 33–52 (2012) - 15.
Burer, S., Letchford, A.: On nonconvex quadratic programming with box constriants. SIAM J. Optim.

**20**(2), 1073–1089 (2009) - 16.
Burer, S., Monteiro, R., Choi, C.: SDPLR 1.03-beta user’s guide (short version) (2009). http://sburer.github.io/files/SDPLR-1.03-beta-usrguide.pdf

- 17.
Burer, S., Vandenbussche, D.: Globally solving box-constrained nonconvex quadratic programs with semdefinite-based finite branch-and-bound. Comput. Optim. Appl.

**43**, 181–195 (2009) - 18.
Caprara, A., Fischetti, M.: \({\{0, \frac{1}{2}\}}\) chvátal-gomory cuts. Math. Program.

**74**, 221–235 (1996) - 19.
Chvátal, V.: Edmonds polytopes and weakly Hamiltonian graphs. Math. Program.

**5**, 29–40 (1973) - 20.
Dolan, E., Moré, J.: Benchmarking optimization software with performance profiles. Math. Program.

**91**, 201–213 (2002) - 21.
Dong, H.: Relaxing nonconvex quadratic functions by multiple adaptive diagonal perturbations. SIAM J Optim

**26**(3), 1962–1985 (2014). https://doi.org/10.1137/140960657 - 22.
Dong, H., Linderoth, J.: On valid inequalities for quadratic programming with continuous variables and binary indicators. In: IPCO 2013: The Sixteenth Conference on Integer Programming and Combinatorial Optimization, vol. 7801, pp. 169–180. Springer (2013)

- 23.
Gomory, R.E.: Outline of an algorithm for integer solutions to linear programs. Bull. Am. Math. Mon.

**64**, 275–278 (1958) - 24.
Hansen, P., Jaumard, B., Ruiz, M., Xiong, J.: Global minimization of indefinite quadratic functions subject to box constraints. Naval Res. Logist.

**40**(3), 373–392 (1993) - 25.
Horst, H., Pardalos, P.M., Thoai, V.: Introduction to Global Optimization, 2nd edn. Kluwer, Dordrecht (2000)

- 26.
Koster, A., Zymolka, A., Kutschka, M.: Algorithms to separate 0,1/2-Chvátal–Gomory cuts. Algorithmica

**55**(2), 375–391 (2009) - 27.
McCormick, G.P.: Computability of global solutions to factorable nonconvex programs: part I—convex underestimating problems. Math. Program.

**10**, 147–175 (1976) - 28.
Misener, R., Smadbeck, J.B., Floudas, C.A.: Dynamically-generated cutting planes for mixed-integer quadratically-constrained quadratic programs and their incorporation into GloMIQO 2.0. Optim. Methods Softw.

**30**, 215–249 (2015) - 29.
Padberg, M.: The boolean quadric polytope: some characterics, facets, and relatives. Math. Program.

**45**, 139–172 (1989) - 30.
Padberg, M.W.: Total unimodularity and the Euler-subgraph problem. Oper. Res. Lett.

**7**(4), 173–179 (1988) - 31.
Saxena, A., Bonami, P., Lee, J.: Convex relaxations of non-convex mixed integer quadratically constrained programs: projected formulations. Math. Program.

**130**, 359–413 (2011). Version with appendix available at http://www.optimization-online.org/DB_FILE/2008/11/2145.pdf - 32.
Sherali, H., Tuncbilek, C.: A new reformulation-convexification approach for solving nonconvex quadratic programming problems. J. Glob. Optim.

**7**, 1–31 (1995) - 33.
Sherali HD, Alameddine AR (1990) An explicit characterization of the convex envelope of a bivariate function over special polytopes. Ann. Oper. Res. Comput. Methods Glob. Optim.

**25**(1): 197–210 - 34.
Shor, N.Z.: Quadratic optimization problems. Sov. J. Circuits Syst. Sci.

**25**(6), 1–11 (1987) - 35.
Simone, C.D.: The cut polytope and the boolean quadric polytope. Discrete Math.

**79**, 71–75 (1989) - 36.
Sturm, J.F.: Using SeDuMi 1.02, a MATLAB toolbox for optimization over symmetric cones. Optim. Methods Softw.

**11–12**, 625–653 (1999) - 37.
The MOSEK command line tool. Version 7.1 (revision 51) (2016). http://docs.mosek.com/7.1/tools/index.html

- 38.
Tawarmalani, M., Sahinidis, N.: A polyhedral branch-and-cut approach to global optimization. Math. Program.

**103**, 225–249 (2005) - 39.
Tawarmalani, M., Sahinidis, N.V.: Global optimization of mixed integer nonlinear programs: a theoretical and computational study. Math. Program.

**99**, 563–591 (2004) - 40.
Vandenbussche, D., Nemhauser, G.L.: A branch-and-cut algorithm for nonconvex quadratic programs with box constraints. Math. Program.

**102**, 559–575 (2005) - 41.
Yajima, Y., Fujie, T.: A polyhedral approach for nonconvex quadratic programming problems with box constraints. J. Glob. Optim.

**13**, 151–170 (1998)

## Author information

### Affiliations

### Corresponding author

## Additional information

The work of author Linderoth is supported in part by the U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research, Applied Mathematics program under Contract Number AC02-06CH11357. IBM Research is also gratefully acknowledged for creating the vibrant research environment where most of this work was accomplished.

## Appendix

### Appendix

### Transforming valid inequalities

We next describe how to transform valid inequalities for \({ \textsc {bqp}}\) to be valid inequalities for the set \(\mathbf {conv}(\mathcal {B}(\ell ,u)) =: { \textsc {bqp}}(\ell ,u)\) so that they can be applied to box-constrained quadratic programs with arbitrary bounds. We explain the transformations in the context of finding a valid inequality violated by a point \((\hat{x},\hat{X})\) with variable bounds \(\ell _i \le x_i \le u_i \ \forall i \in N\). The first step is to transform the variables via the equations

The transformation (21) is obtained by scaling each variable in the product

and replacing \(x_i x_j\) with \(X_{ij}\). The point \((x,X) \in { \textsc {bqp}}(\ell ,u)\) if and only if the transformed point \((x^\prime ,X^\prime ) \in { \textsc {bqp}}\). The point \((x^\prime ,X^\prime )\) is given to CPLEX’s standard \(0\)-\(\frac{1}{2}\) cut generator, which attempts to separate \((x^\prime ,X^\prime )\) for \({ \textsc {bqp}}\). If successful, a “canonical,” inequality of the form

is returned. We undo the transformation by substituting (20) and (21) into (22) to get the following equivalent inequality for \({ \textsc {bqp}}(\ell ,u)\):

To implement inequality (23), the coefficient of the \(x_i\) variable in (22) is multiplied by \((u_i-\ell _i)^{-1}\) and the quantity \(\sum _{j : \{i,j\}\in E} \ B_{ij} \ell _j /(u_i-\ell _i)(u_j-\ell _j)\) is subtracted from the coefficient. For each \(X_{ij}\) variable, we scale the coefficient by coefficients of the \(X_{ij}\) variables by \((u_i - \ell _i)(u_j - \ell _j)\). We also add the quantity

to the right-hand-side of the returned inequality. When computing the coefficient of each variable, if the difference in bounds for the variable is smaller than a pre-specified tolerance, (\(u_i - \ell _i < \epsilon \)), we take \(u_i = \ell _i + 1\) when calculating the transformations (20) (21) and (23). The resulting inequality is clearly valid, since it is valid for a relaxation of the nodal subproblem being considered. Further, numerical difficulties arising from possibly large coefficient values in (23) are avoided.

When these transformations are applied to the defining inequalities of the Boolean Quadric Polytope, one obtains the well-known general version of the McCormick inequalities:

### bqp Bounds

Table 8 contains the raw numbers that were aggregated to make Tables 1 and 2. The value in the column \(z_{BoxQP\ }\) is the optimal solution value, \(z_{{ \textsc {bqp}}}^U\) is the upper bound on \(z_{BoxQP\ }\) obtained by the bqprestriction, and \(z_{{ \textsc {bqp}}}^L\) is the lower bound on \(z_{BoxQP\ }\) obtained from the bqprelaxation. The \(z_{{ \textsc {bqp}}}^L\)(root) column contains the final root relaxation value from CPLEX when solving the integer program associated with \(z_{{ \textsc {bqp}}}^L\), and \(z_{\Delta ^+}\) is the bound at the root that could be obtained if exact separation of the odd cycle inequalities was done. The column \(z_{\mathcal {M}}\) contains the lower bound obtained from optimizing over the McCormick relaxation \(\mathcal {M}\).

### Algorithmic improvements

Table 9 contains the raw numbers that were aggregated to make Table 3. The value in the column \(z_{BoxQP\ }\) is the optimal solution value, \(z_{\mathcal {M}^2}+\) \(0\)-\(\frac{1}{2}\) is the bound obtained from the strengthened McCormick relaxation with CPLEX \(0\)-\(\frac{1}{2}\)-cuts added, \(z_{\mathcal {M}}+\) \(0\)-\(\frac{1}{2}\)the bound from the standard McCormick relaxation with CPLEX \(0\)-\(\frac{1}{2}\)-cuts added. The columns \(z_{\mathcal {M}^2}\) and \(z_{\mathcal {M}}\) show the lower bounds obtained from the strengthened and traditional McCormick relaxation, respectively.

Table 10 contains raw numbers on the computational performance of algorithmic techniques described in Sects. 4.2 and 4.3.

### SDP relaxation bounds

This appendix contains computational time and relaxation bound information for various SDP-based approaches for solving BoxQP , as described in Sect. 5.1. For the results of the two methods SBL and Dong, the reader is invited to refer directly to the papers [21, 31] (Table 11).

### Solver comparison

This appendix contains the raw performance numbers comparing different solvers for BoxQP . The experiment is described in Sect. 5.2 (Table 12).

### CPLEX performance

This appendix contains the raw performance numbers for the computational experiment comparing the performance of CPLEX v12.6.3 with and without the BQP-based cutting planes described in Sect. 5.3 (Table 13).

## Rights and permissions

## About this article

### Cite this article

Bonami, P., Günlük, O. & Linderoth, J. Globally solving nonconvex quadratic programming problems with box constraints via integer programming methods.
*Math. Prog. Comp.* **10, **333–382 (2018). https://doi.org/10.1007/s12532-018-0133-x

Received:

Accepted:

Published:

Issue Date:

### Keywords

- Nonconvex quadratic programming
- Global optimization
- Boolean Quadric Polytope

### Mathematics Subject Classification

- 90C20
- 90C26
- 90C57