Q01SUBS

Unconstrained quadratic 0-1 programming for both dense and sparse matrices (including concave quadratic problems with box constraints), the maximum clique problem, and random test problem generation for all these problems

Branch and bound with depth-first search and dynamic variable selection. Research reports providing the theoretical basis for the algorithms are available on request.

Software is written in Fortran using integer arithmetic in all cases. The code is documented by comment statements. User feedback is much appreciated.

Need more info?

Contact:

Professor Panos Pardalos 
303 Weil Hall 
Department of Industrial Engineering 
University of Florida 
Gainesville, FL 32611
pardalos@math.ufl.edu

References:

P. M. Pardalos, Construction of test problems in quadratic bivalent programming, ACM Trans. Math. Software 17 (1991), pp. 74--87.

P. M. Pardalos and G. P. Rodgers, A branch and bound algorithm for the maximum clique problem, Comp. and Oper. Research (1991, to appear).

P. M. Pardalos and G. P. Rodgers, Computing aspects of a branch and bound algorithm for quadratic zero-one programming, Computing 45 (1990), pp. 131--144.

P. M. Pardalos and G. P. Rodgers, Parallel branch and bound algorithms for quadratic zero-one programs on the hypercube architecture, Ann. Oper. Res. 22 (1990), pp. 271--292.

[ Optimization Software Guide | OTC Home Page | NEOS Server | NEOS Guide ]