Finding Membership Functions for Bioinformatics Data

Mendel 2011 : 17th International Conference on Soft Computing : Evolutionary Computation, Genetic Programming, Fuzzy Logic, Rough Sets, Neural Networks, Fractals, Bayesian Methods 2011
Madara Gasparoviča-Asīte, Ludmila Aleksejeva, Irēna Tuleiko

This paper studies various fuzzy membership function construction methods to find the most appropriate ones that can be used in bioinformatics data analysis. Correct construction of membership functions increases accuracy of classification because the obtained results are strictly related to classification. In this research four membership function methods have been analyzed by their principle of working, performance and complexity. For practical experiments real data sets have been used - Gastric cancer, Golub leukemia, Singh prostate and Leukemia data sets. In order to check data sets and membership functions computed with different methods Fuzzy Prism algorithm was used. As a result, conclusions were made about the most appropriate method of constructing membership functions in bioinformatics data.

Atslēgas vārdi
Membership function; Bioinformatics data, Classification; Gastric cancer; Lymphoma data; Fuzzy IFTHEN rules

Gasparoviča-Asīte, M., Aleksejeva, L., Tuleiko, I. Finding Membership Functions for Bioinformatics Data. No: Mendel 2011 : 17th International Conference on Soft Computing : Evolutionary Computation, Genetic Programming, Fuzzy Logic, Rough Sets, Neural Networks, Fractals, Bayesian Methods, Čehija, Brno, 15.-17. jūnijs, 2011. Brno: Brno University of Technology, 2011, 133.-140.lpp. ISBN 978-80-2144302-0.

Publikācijas valoda
English (en)