The Computational Intelligence Research Group (CIRG) is based at the University of Pretoria, in the Department of Computer Science, and focuses on research in the broad realm of computational intelligence. Under this umbrella, work is done in the fields of swarm intelligence, evolutionary computation, neural networks, and artificial immune systems, with applications to optimization, classification, predication, data and text mining, data clustering, image analysis, bioinformatics, and financial analysis. A major application area is in optimization with a focus on multi-objective optimization, optimization in dynamic environments, and locating multiple solutions to highly multi-modal problems. Work is also done in classification, predication, clustering, and data anslsysis in dynamic environments.
We conduct theoretical research, focusing on the development of new and improvement of existing computational intelligence algorithms. We also do application-oriented research, using computational intelligence techniques to solve real-world problems. We are also actively involved with industry through consultation and contract research. Based on these focuses, CIRG thus strives to fulfill the following twofold mission:
- To deliver graduates of exceptional standard with a very good knowledge of modern computational intelligence;
- To apply modern computational intelligence techniques to effectively solve complex real-world problems.
CIRG had its beginnings in 1998, when Professor Andries Engelbrecht joined the Computer Science Department at the University of Pretoria. While at first consisting of only two postgraduate students, and not recognised as an official research group, student numbers quickly increased due to the exciting nature of its work. By 2003 CIRG had been officially established as a research group, and lab space was allocated for its quickly growing complement of postgraduate students.
Currently, CIRG is still headed by Professor Engelbrecht, and consists of a further five permanent staff members, and approximately 40 postgraduate students on Masters and Doctoral level. In addition, several undergraduate students (on Honours B.Sc level) are also completing projects within CIRG. We also offer and manage year-projects to a selection of undergraduate students on third year level.
The major thrust of the entire research group is currently towards the continued development of CIlib: A collaborative, open source framework for the implementation of a vast number of CI algorithms and their hybridisations. The drive of this initiative is towards a single repository of implementations with verified accuracy and quality.
CIRG has recently launched an exciting project to install an IBM cloud computing solution, for the purposes of running CIlib-based simulations in a highly distributed fashion, as well as to facilitate the capturing of experimental results and statistics. It is hoped that this project will not only accelerate CIRG's own research, but ultimately foster better collaboration and sharing of results and resources across the entire CI research community. The final outcome will be a repository of optimization benchmark data with summarative statistics.