The Levy Particle Swarm

Dr Toby Riche

Goldsmiths College, University of London

Date and time: 11.30am - 12.30pm, Friday 30 June, 2006

Venue: 10.08.04 (Building 10, Level 8, Room 4)

Abstract:

Many foragers and wandering animals have been shown to follow a Lévy distribution of steps and it is conjectured here that this distribution may be useful for optimization algorithms too. This paper investigates the effectiveness of replacing the particle dynamics within Particle Swarm Optimization (PSO) by random sampling from a Lévy distribution. The resulting Lévy PSO is demonstrated to perform as well, or better, than a standard PSO or equivalent Gaussian models over a range of benchmark problems.

About the speaker:

Toby Richer received his Bachelor of Engineering (Computer Systems) and Bachelor of Science (Mathematical Science) from the University of Adelaide in 2000. He completed his doctoral dissertation, "Dynamic Territoriality for Multi-Robot Systems", at the University of South Australia in 2005. Since then, he has been employed as a research officer at Goldsmiths College, University of London. His general areas of interest are swarm intelligence and biologically-inspired robotics; he is currently researching alternative formulations and metrics for Particle Swarm Optimisation.

Seminar Organisation

Seminars are free and open to the general public. No booking is necessary. If you are interested in giving a presentation in this seminar series, or to make suggestions for speakers, please contact Xiaodong Li, the seminar co-ordinator.