The constant objective value property for multidimensional assignment problems

An instance of a combinatorial optimization problem is said to have the constant objective value property (COVP) if every feasible solution has the same objective function value. In this paper our goal is to characterize the set of all instances with the COVP for multidimensional assignment problems...

Full description

Saved in:
Bibliographic Details
Published inDiscrete optimization Vol. 19; pp. 23 - 35
Main Authors Ćustić, Ante, Klinz, Bettina
Format Journal Article
LanguageEnglish
Published Elsevier B.V 01.02.2016
Subjects
Admissible transformation
Constant objective value
Multidimensional assignment problem
Sum-decomposable array
Admissible transformation
Sum-decomposable array
Multidimensional assignment problem
Constant objective value
Online AccessGet full text

Cover

More Information
Summary:An instance of a combinatorial optimization problem is said to have the constant objective value property (COVP) if every feasible solution has the same objective function value. In this paper our goal is to characterize the set of all instances with the COVP for multidimensional assignment problems. Our central result deals with planar d-dimensional assignment problems. We show that such constant objective value instances are characterized by so-called sum-decomposable arrays with appropriate parameters. This adds to the known result for the axial d-dimensional case.
ISSN:1572-5286
1873-636X
DOI:10.1016/j.disopt.2016.01.004