New methods to color the vertices of a graph
Communications of the ACM
Cliques, Coloring, and Satisfiability: Second DIMACS Implementation Challenge, Workshop, October 11-13, 1993
A New Genetic Local Search Algorithm for Graph Coloring
PPSN V Proceedings of the 5th International Conference on Parallel Problem Solving from Nature
A graph coloring heuristic using partial solutions and a reactive tabu scheme
Computers and Operations Research
Diversity Control and Multi-Parent Recombination for Evolutionary Graph Coloring Algorithms
EvoCOP '09 Proceedings of the 9th European Conference on Evolutionary Computation in Combinatorial Optimization
Bumblebees: a multiagent combinatorial optimization algorithm inspired by social insect behaviour
Proceedings of the first ACM/SIGEVO Summit on Genetic and Evolutionary Computation
Adaptation of a multiagent evolutionary algorithm to NK landscapes
Proceedings of the 15th annual conference companion on Genetic and evolutionary computation
| Hi-index | 0.00 |
In this paper, two new metaheuristic algorithms for the graph coloring problem are introduced. The first one is a population-based multiagent evolutionary algorithm (MEA), using a multiagent system, where an agent represents a tabu search procedure. Rather than using a single long-term local search procedure, it uses more agents representing short term local search procedures. Instead of a specific crossover, MEA uses relatively general mechanisms from artificial life, such as lifespans and elite list [3, 4]. We are introducing and investigating a new parametrization system, along with a mechanism of reward and punishment for agents according to change in their fitness. The second algorithm is a pseudo-reactive tabu search (PRTS), introducing a new online learning strategy to balance its own parameter settings. Basically, it is inspired by the idea to learn tabu tenure parameters instead of using constants. Both algorithms empirically outperform basic tabu search algorithm TabuCol [8] on the well-established DIMACS instances [10]. However, they achieve this by using different strategies. This indeed shows a difference in potential of population-based and learning-based graph coloring metaheuristics.