Is there a computational advantage to representing evaporation rate in ant colony optimization as a gaussian random variable?

Is there a computational advantage to representing evaporation rate in ant colony optimization as a gaussian random variable?

Files

Department

Computer Science & Engineering Department

Abstract

We propose an ACO (Ant Colony Optimization) variation in which the evaporation rate, instead of being constant as is common in standard ACO algorithms, is a Gaussian random variable with non-negligible variance. In experimental results in the context of MAX-MIN Ant System (MMAS) and the Traveling Salesman Problem (TSP), we find that our variation performs considerably better than MMAS when the number of iterations is small, and that its performance is slightly better than MMAS when the number of iterations is large. © 2012 ACM.

Publication Date

8-13-2012

Document Type

Book Chapter

Book Title

GECCO'12 - Proceedings of the 14th International Conference on Genetic and Evolutionary Computation

Editors

Terence Soule

ISBN

9781450311779

Publisher

Association for Computing Machinery

City

New York, NY

First Page

1

Last Page

8

Keywords

ant colony optimization, pheromone decay

Is there a computational advantage to representing evaporation rate in ant colony optimization as a gaussian random variable?

Share

COinS