Exact solution approaches for integer linear generalized maximum multiplicative programs through the lens of multi-objective optimization

We study a class of single-objective non-linear optimization problems, the so-called Integer Linear Generalized Maximum Multiplicative Programs (IL-GMMP). This class of optimization problems has a significant number of applications in different fields of study including but not limited to game theor...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Computers & operations research Jg. 137; S. 105549
Hauptverfasser: Saghand, Payman Ghasemi, Charkhgard, Hadi
Format: Journal Article
Sprache:Englisch
Veröffentlicht: New York Elsevier Ltd 01.01.2022
Pergamon Press Inc
Schlagworte:
Algorithms
Exact solutions
Game theory
Generalized maximum multiplicative programming
Geometric mean optimization
Integer programming
Linear programming
Mathematical analysis
Mixed integer
Multi-objective integer linear programming
Multiple objective analysis
Nash bargaining solution
Nash social welfare
Operations research
Optimization
Pareto optimization
Solvers
System reliability
Nash social welfare
Generalized maximum multiplicative programming
Multi-objective integer linear programming
Nash bargaining solution
Geometric mean optimization
ISSN:0305-0548, 1873-765X, 0305-0548
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:We study a class of single-objective non-linear optimization problems, the so-called Integer Linear Generalized Maximum Multiplicative Programs (IL-GMMP). This class of optimization problems has a significant number of applications in different fields of study including but not limited to game theory, systems reliability, and conservation planning. An IL-GMMP can be reformulated as a mixed integer Second-Order Cone Program (SOCP) and therefore, can be solved effectively by commercial solvers such as IBM ILOG CPLEX, Gurobi, and FICO Xpress. In this study, we show that IL-GMMPs can be viewed as special cases of the problem of optimization over the efficient (or Pareto-optimal) set in multi-objective integer linear programming. Based on this observation, we develop three exact solution approaches with a desirable property: they only solve a finite number of single-objective integer linear programs to compute an optimal solution of an IL-GMMP (which is nonlinear). Through an extensive computational study with 57600 experiments, we compare the performance of all three algorithms using the three main commercial single-objective integer linear programming solvers in the market: CPLEX, Gurobi, and Xpress. We also compare the performance of our algorithms using the mixed integer SOCP solvers of CPLEX, Gurobi, and Xpress. The results show that the choice of a commercial solver impacts the solution time dramatically and that, by the right choice of solver, one of our proposed algorithms is significantly faster than other methods. We also illustrate that although it is possible to linearize IL-GMMPs, commercial solvers struggle to solve such linearized instances. •We study a class of single-objective integer nonlinear problems.•We develop three multi-objective based algorithms for solving this class of problems.•The proposed algorithms lie in criterion space search algorithms.•The proposed algorithms employ novel cut-generating and bounding mechanisms.•We show the efficacy of the proposed algorithms using a computational study.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0305-0548
1873-765X
0305-0548
DOI:10.1016/j.cor.2021.105549