Evolutionary approach to combined multiple models tuning

Issue title: Extended papers selected from KES-2006

Article type: Research Article

Authors: Stiglic, Gregor^{; *} | Kokol, Peter

Affiliations: Laboratory for System Design, Faculty of Electrical Engineering and Computer Science, Smetanova 17, 2000 Maribor, Slovenia

Correspondence: [*] Corresponding author. Tel.: +386 2 220 7465; Fax: +386 2 220 7272; E-mail: gregor.stiglic@uni-mb.si

Abstract: This paper presents a combination of Combined Multiple Models (CMM) technique and evolutionary approach that is used for tuning of multiple parameters. Proposed hybrid classifier was tested in microarray gene expression analysis domain. This domain was chosen intentionally, because of the nature of Combined Multiple Models classifiers that are specialized in solving problems with high dimensionality and contain low number of samples. Evolutionary tuning of parameters in combination with validation dataset enables fine tuning of parameters that are usually set to pre-defined values. Using this technique another step in leveling the accuracy of comprehensible classifiers to those represented by ensembles of classifiers was made.

DOI: 10.3233/KES-2007-11405

Journal: International Journal of Knowledge-based and Intelligent Engineering Systems, vol. 11, no. 4, pp. 227-235, 2007