The weighted independent domination problem: Integer linear programming models and metaheuristic approaches

•We tackle the weighted independent domination problem.•Three different integer linear programming models are presented.•Two greedy heuristics are studied.•A population-based iterated greedy algorithm is developed.•This algorithm is also applied in a framework based on instance reduction. This work...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:European journal of operational research Ročník 265; číslo 3; s. 860 - 871
Hlavní autori: Pinacho Davidson, Pedro, Blum, Christian, Lozano, Jose A.
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Elsevier B.V 16.03.2018
Predmet:
Combinatorial optimization
Construct, merge, solve & adapt
Heuristics
Integer linear programming
Population-based iterated greedy
Integer linear programming
Combinatorial optimization
Heuristics
Construct, merge, solve & adapt
Population-based iterated greedy
ISSN:0377-2217, 1872-6860
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:•We tackle the weighted independent domination problem.•Three different integer linear programming models are presented.•Two greedy heuristics are studied.•A population-based iterated greedy algorithm is developed.•This algorithm is also applied in a framework based on instance reduction. This work deals with the so-called weighted independent domination problem, which is an NP-hard combinatorial optimization problem in graphs. In contrast to previous work, this paper considers the problem from a non-theoretical perspective. The first contribution consists in the development of three integer linear programming models. Second, two greedy heuristics are proposed. Finally, the last contribution is a population-based iterated greedy metaheuristic which is applied in two different ways: (1) the metaheuristic is applied directly to each problem instance, and (2) the metaheuristic is applied at each iteration of a higher-level framework – known as construct, merge, solve and adapt – to sub-instances of the tackled problem instances. The results of the considered algorithmic approaches show that integer linear programming approaches can only compete with the developed metaheuristics in the context of graphs with up to 100 nodes. When larger graphs are concerned, the application of the populated-based iterated greedy algorithm within the higher-level framework works generally best. The experimental evaluation considers graphs of different types, sizes, densities, and ways of generating the node and edge weights.
ISSN:0377-2217
1872-6860
DOI:10.1016/j.ejor.2017.08.044