From Theory to Practice in Particle Swarm Optimization

Abstract

The purpose of this chapter is to draw attention to two points that are not always well understood, namely, a) the ``balance'' between exploitation and exploration may be not what we intuitively think, and b) a mean best result may be meaningless. The second point is obviously quite important when two algorithms are compared. These are discussed in the appendix. We believe that these points would be useful to researchers in the field for analysis and comparison of algorithms in a better and rigorous way, and help them design new powerful tools.