Logo

CIRG - Research  -  Swarm Intelligence 



[Bullet] Home
About
News
[Bullet]

- NN
- DM
- SI
- EC
- MAS
- AIS
- IA
- Bioinf
- Games
- Opt
- FA
- Industry

Publications
People
Resources
Links
Contact Us

OVERVIEW

The Swarm Intelligence focus area is currently the most active in the group, with the largest number of members. The focus area's main interest is particle swarm optimization (PSO), with the development of new and improved PSO algorithms. Theoretical analyses of PSO are also being done, with convergence proofs being studied. Techniques are developed for constrained optimization, niching (locating multiple solutions), multi-objective optimization, dynamic optimization problems, and to cope with discrete search spaces.

Applications of PSO techniques that are under investigation include the coevolutionary training of neural networks for game playing and financial traders, scheduling, image analysis, and data clustering. The research focus area is also investigating the application of ant colony optimization techniques to exploratory data analysis, workload distribution in computer grids, energy efficient routing in mobile ad hoc networks, and network topology design.

ACTIVE MEMBERS

List the current members actively doing research in this focus area. [ Show ]

ALUMNI MEMBERS

L Schoeman

PhD Completed in 2010

J Grobler

PhD Started in 2009
M.Eng Completed in 2009

S Khan

PhD Completed in 2009

D Barla-Szabo

M.Sc Started in 2003
Hons-B.Sc Completed in 2002

M Neethling

M.Sc Completed in 2008

F Zablocki

M.Sc Completed in 2008

E Papacostantis

M.Sc Started in 2004
Hons-B.Sc Completed in 2003

L Messerschmidt

M.Sc Completed

C Naicker

PhD Started in 2007
M.Sc Completed in 2006
Hons-B.Sc Completed in 2002

J Conradie

Hons-B.Sc Completed in 2004

J du Plessis

M.Sc Completed in 2005

E Peer

M.Sc Completed in 2005

M Omran

PhD Completed in 2005

G Pampara

M.Sc Started in 2005
Hons-B.Sc Completed in 2004

E van Loggerenberg

M.Sc Started in 2005
Hons-B.Sc Completed in 2004

N Franken

PhD Started in 2005
M.Sc Completed in 2004
Hons-B.Sc Completed in 2002

R Brits

M.Sc Completed in 2003
Hons-B.Sc Completed in 2000

U Paquet

M.Sc Completed in 2003

F van den Bergh

PhD Completed in 2002

A Ismail

PhD Started in 2005
M.Sc Completed in 2001

GROUP PUBLICATIONS

List publications of this research focus area. [ Show ]

MEMBER PROFILE



 Name:

 Nelis Franken

Portrait photo

 E-mail:

 nfranken@cs.up.ac.za

 Group(s):

 Evolutionary Computation
 Swarm Intelligence
 Games

 

 Degree specific information: PhD

 Title:

 Variable Length Particles for PSO

 Abstract:

Not available

 Supervisor / Co-Supervisor:

 AP Engelbrecht

 Thesis:

 Not available for download yet.

 

 Degree specific information: M.Sc

 Title:

 PSO-Based Coevolutionary Game Learning

 Abstract:

Games have been investigated as computationally complex problems since the inception of artificial intelligence in the 1950's. Originally, search-based techniques were applied to create a competent (and sometimes even expert) game player. The search-based techniques, such as game trees, made use of human-defined knowledge to evaluate the current game state and recommend the best move to make next. Recent research has shown that neural networks can be evolved as game state evaluators, thereby removing the human intelligence factor completely. This study builds on the initial research that made use of evolutionary programming to evolve neural networks in the game learning domain. Particle Swarm Optimisation (PSO) is applied inside a coevolutionary training environment to evolve the weights of the neural network. The training technique is applied to both the zero sum and non-zero sum game domains, with specific application to Tic-Tac-Toe, Checkers and the Iterated Prisoner's Dilemma (IPD). The influence of the various PSO parameters on playing performance are experimentally examined, and the overall performance of three different neighbourhood information sharing structures compared. A new coevolutionary scoring scheme and particle dispersement operator are defined, inspired by Formula One Grand Prix racing. Finally, the PSO is applied in three novel ways to evolve strategies for the IPD -- the first application of its kind in the PSO field. The PSO-based coevolutionary learning technique described and examined in this study shows promise in evolving intelligent evaluators for the aforementioned games, and further study will be conducted to analyse its scalability to larger search spaces and games of varying complexity.

 Supervisor / Co-Supervisor:

 AP Engelbrecht

 Thesis:

 Download




You are visitor #117518
Contact webmaster
Back to top

QualNet Network Simulator University Program Valid XHTML 1.0! Valid CSS!


Computational Intelligence Research Group
University of Pretoria
Copyright © 2017