Сomparison of learning criteria for fuzzy classifier with voting rules

  • Serhiy Shtovba Vinnytsia National Technical University
  • Anastasiia Galushchak Vinnytsia National Technical University
Keywords: classification, fuzzy knowledge base, learning, voting rules, learning criteria, main competitors

Abstract

In fuzzy classifiers decision-making is based on linguistic rules <If - then>, antecedents of which contain fuzzy terms “low”, “average”, “high” etc. To increase the correctness fuzzy classifier is learned by experimental data. We study a fuzzy classifier with voting rules in which by the result of logic inference is class with maximum total supports by all the rules. New criteria of fuzzy classifier learning are suggested, they take into account the difference of memberships of fuzzy inference only to main competitors. In case of correct classification, main competitor of the taken decision is the class with the second membership degree. In case of incorrect classification, decision, taken by mistake is main competitor of the correct class. Computer experiments, dealing with learning of fuzzy classifier for UCI-problem of Italian wines recognition proved significant advantage of new learning criteria.

Author Biographies

Serhiy Shtovba, Vinnytsia National Technical University
Doctor of Sc(Eng.)., Professor, Department of Computer Control Systems
Anastasiia Galushchak, Vinnytsia National Technical University
Assistant, Department of Computer Control Systems

References

1. Kuncheva L. I. Fuzzy classifier design, Studies in Fuzziness and Soft Computing / L. I. Kuncheva . – Vol. 49. – Berlin – Heidelberg: Springer-Verlag, 2000. – 314 p.
2. Rotshtein A. N. Design and tuning of fuzzy rule-based system for medical diagnosis / A. N. Rotshtein,
N. H. Teodorescu, , A. Kandel, L. C. Jain // Fuzzy and Neuro-Fuzzy Systems in Medicine. – Boca–Raton : CRC–Press, 1998. P. 243 – 289.
3. Ishibuchi H. Voting in fuzzy rule-based systems for pattern classification problems /
H. Ishibuchi, T. Nakashima, T. Morisawa // Fuzzy Sets and Systems. – 1999. – Vol. 103, №2. – P. 223 – 238.
4. Ishibuchi H. Classification and modeling with linguistic information granules: advanced approaches advanced approaches to linguistic data mining / H. Ishibuchi, T. Nakashima, M. Nii. – Berlin – Heidelberg: Springer-Verlag, 2005. – 307 p.
5. Shtovba S. Tuning the fuzzy classification models with various learning criteria: the case of credit data classification / S. Shtovba, O. Pankevich, G. Dounias // Proc. of Inter. Conference on Fuzzy Sets and Soft Computing in Economics and Finance. St. Petersburg (Russia). St. Petersburg: Russian Fuzzy Systems Association. – 2004. – Vol. 1. –P. 103 – 110.
6. Штовба С. Д. Проектирование нечетких систем средствами MATLAB / С. Д. Штовба. – М. : Горячая линия – Телеком, 2007. – 288 с.
7. Штовба С. Д. Порівняння критеріїв навчання нечіткого класифікатора / С. Д. Штовба // Вісник Вінницького політехнічного інституту. – 2007. – № 6. – С. 84 – 91.
8. Штовба С. Д. Анализ критериев обучения нечеткого классификатора / С. Д. Штовба,
О. Д. Панкевич, А. В. Нагорна // Автоматика и вычислительная техника. 2015. № 3. С. 5 16.
9. Панкевич О. Д. Діагностування тріщин будівельних конструкцій за допомогою нечітких баз знань. Монографія / О. Д. Панкевич, С. Д. Штовба. – Вінниця: УНІВЕРСУМ – Вінниця, 2005. – 108 с.
10. Штовба С. Д. Обеспечение точности и прозрачности нечеткой модели Мамдани при обучении по экспериментальным данным / С. Д. Штовба // Проблемы управления и информатики. – 2007. – № 4. – С. 102 – 114.
Published
2016-03-02
How to Cite
[1]
S. Shtovba and A. Galushchak, “Сomparison of learning criteria for fuzzy classifier with voting rules”, SWVNTU, no. 4, Mar. 2016.
Section
Information Technologies and Computer Engineering