Meta-Learning and Model Selection in Multiobjective Evolutionary Algorithms

0384809 - ÚI 2013 RIV US eng C - Konferenční příspěvek (zahraniční konf.)
Pilát, M. - Neruda, Roman
Meta-Learning and Model Selection in Multiobjective Evolutionary Algorithms.
Proceedings 2012 11th International Conference on Machine Learning and Applications ICMLA 2012. Los Alamitos: IEEE Computer Society, 2012 - (Wani, M.; Khoshgoftaar, T.; Zhu, X.; Seliya, N.), s. 433-438. ISBN 978-1-4673-4651-1.
[ICMLA 2012. International Conference on Machine Learning and Applications /11./. Boca Raton (US), 12.12.2012-15.12.2012]
Grant CEP: GA ČR GAP202/11/1368; GA ČR GD201/09/H057
Grant ostatní: UK(CZ) SVV-265314
Institucionální podpora: RVO:67985807
Klíčová slova: multiobjective optimization * surrogate modelling * meta-learning * model selection
Kód oboru RIV: IN - Informatika

Most existing surrogate based evolutionary algorithms deal with only one model selected by the authors and different models are not considered. In this paper we propose a framework which enables automatic selection of types of surrogate models, and evaluate the effect of the type of selection on the overall performance of the resulting evolutionary algorithm. Two different types of model selection are tested and compared both in pre-selection scenario and in local search scenario.
Trvalý link: http://hdl.handle.net/11104/0214327