Extended trail reinforcement strategies for ant colony optimization

Authors:
Nikola Ivkovic;Mirko Malekovic;Marin Golub
Affiliations:
Faculty of Organization and Informatics, University of Zagreb, Croatia;Faculty of Organization and Informatics, University of Zagreb, Croatia;Faculty of Electrical Engineering and Computing, University of Zagreb, Croatia
Venue:
SEMCCO'11 Proceedings of the Second international conference on Swarm, Evolutionary, and Memetic Computing - Volume Part I
Year:
2011

Citing 4
Cited 0

Ant algorithms for discrete optimization

Artificial Life
MAX-MIN Ant system

Future Generation Computer Systems
Ant colony system: a cooperative learning approach to the traveling salesman problem

IEEE Transactions on Evolutionary Computation
Ant system: optimization by a colony of cooperating agents

IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics

Quantified Score

Hi-index	0.00

Visualization

Abstract

Ant colony optimization (ACO) is a metaheuristic inspired by the foraging behavior of biological ants that was successfully applied for solving computationally hard problems. The fundamental idea that drives the ACO is the usage of pheromone trails for accumulating experience about the problem that is been solved. The best performing ACO algorithms typically use one, in some sense "the best", solution to reinforce trail components. Two main trail reinforcement strategies are used in ACO algorithms: iteration best and global best strategy. This paper extends the reinforcement strategies by using the information from an arbitrary number of previous iterations of the algorithm. The influence of proposed strategies on algorithmic behavior is analyzed on different classes of optimization problems. The conducted experiments showed that using the proposed strategies can improve the algorithm's performance. To compare the strategies we use the Mann---Whitney and Kruskal --- Wallis statistical tests.