May the best optimisation algorithm stand up?
We develop models for metaheuristic research which recognises the need to match algorithms to problems. Information theoretic measures as well as semi-empirical approaches to producing mapping from problems to algorithms are investigated. This mapping, if successful, will encapsulate the knowledge gained from the application of metaheuristics to the spectrum of real problems.
Recent papers in this area of research include:
M Muñoz, M Kirley, S Halgamuge, "A Meta-learning Prediction Model of Algorithm Performance for Continuous Optimization Problems", Parallel Problem Solving from Nature-PPSN XII, 226-235, 2012.
M Munoz, M Kirley, S Halgamuge, "Landscape Characterization of Numerical Optimization Problems Using Biased Scattered Data"
Evolutionary Computation (CEC), 2012 IEEE Congress on
K Steer, A Wirth, S Halgamuge, "Information Theoretic Classification of Problems for Metaheuristics", Simulated Evolution and Learning, 319-328, 2008.
Leader: Saman Halgamuge
Collaborators: Michael Kirley (Computing and Information Systems) , Kent Steer (IBM Research Lab), Xiaodong Li (RMIT University), Marcus Gallagher (University of Queensland)
Sponsors: Australian Research Council
Networks and data in society, Optimisation of resources and infrastructure
artificial intelligence; evolutionary algorithms; optimization