Sensitivity analysis of the bee colony optimization algorithm

Abstract

Bee Colony Optimization (BCO) is a nature-inspired population-based meta-heuristic method that belongs to the class of Swarm intelligence algorithms. BCO was proposed by Lučić and Teodorović, who were among the first to use the basic principles of collective bee intelligence in dealing with optimization problems. Designing a BCO method in principle includes choosing a procedure for constructive/improvement moves, an evaluation function and setting BCO parameters to a suitable values. Topic of this work is addressing the influence of right choice of BCO underlying procedures, such as the choice of loyalty functions, and influence of parameter variations on algorithm performance, by means of visual inspection. Analyses were conducted for simple variant of scheduling problem. Also, to achieve good alternatives for reported solutions, new evaluation methods for scheduling problem are presented.