Hybrid Approach Using Rough Sets and Fuzzy Logic to Pattern Recognition Task

  • Andrzej Zolnierek
  • Marcin Majak
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8073)


In this paper a hybrid classifier construction using rough sets and fuzzy logic is presented. Nowadays, we tackle with many realistic multi-dimensional problems with continuous values and overlaps in the feature space which require sophisticated recognition algorithms. Many methods have been proposed in the literature to improve classification accuracy, but it is increasingly harder to build new classifier from the scratch. Instead, new fusion methods are proposed to overcome this problem. In our rough-fuzzy approach data pre-processing and crisp discretization have a significant impact on the final classification efficiency. To deal with the problem of finding the optimal cuts in the feature space a genetic algorithm was proposed. After the algorithm description, in this paper also simulation investigations using different datasets from UCI Machine Learning Repository are presented.


Rough Sets Fuzzy Logic Hybrid Pattern Recognition Algorithm 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Ishibuchu, H., Yamamoto, T.: Hybridization of fuzzy gbml approaches for pattern classification problem. IEEE Trans. on Systems, Man, and Cybernetics 35(2), 359–365 (2005)CrossRefGoogle Scholar
  2. 2.
    Qinghua, H., Zongxia, X.: Hybrid attribute reduction based on a novel fuzzy-rough model and information granulation. Pattern Recognition 40, 3509–3521 (2007)zbMATHCrossRefGoogle Scholar
  3. 3.
    Hu, Q., Shuang, A.: Robust fuzzy rough classifiers. Fuzzy Sets and Systems 183, 26–43 (2011)MathSciNetzbMATHCrossRefGoogle Scholar
  4. 4.
    Roy, A., Pal, S.: Fuzzy discretization of feature space for a rough set classifier. Pattern Recognition Letters 24(6), 895–902 (2003)zbMATHCrossRefGoogle Scholar
  5. 5.
    Shen, Q., Chouchoulas, A.: A rough-fuzzy approach for generating classifcation rules. Pattern Recognition 35, 2425–2438 (2003)CrossRefGoogle Scholar
  6. 6.
    Wu, Q., Wang, T., Ji-Sheng, L.: New research on fuzzy rough sets. In: Proceedings of the Fifth International Conference on Machine Learning and Cybernetics, pp. 13–16 (2006)Google Scholar
  7. 7.
    Slowinski, R.: Intelligent decision support: handbook of applications and advances of the rough sets theory. Kluwer Academic Publishers, Dordrecht (2010)Google Scholar
  8. 8.
    Pawlak, Z.: Rough sets, decision algorithms and bayes theorem. European Journal of Operational Research 136(1), 181–189 (2002)MathSciNetzbMATHCrossRefGoogle Scholar
  9. 9.
    Khoo, L., Zhai, L.: A prototype genetic algorithm-enhanced rough set-based rule induction system. Computers in Industry 46, 95–106 (2001)CrossRefGoogle Scholar
  10. 10.
    Mendes, R.R.F., Voznika, F.D.B., Freitas, A.A., Nievola, J.C.: Discovering fuzzy classification rules with genetic programming and co-evolution. In: Siebes, A., De Raedt, L. (eds.) PKDD 2001. LNCS (LNAI), vol. 2168, pp. 314–325. Springer, Heidelberg (2001)CrossRefGoogle Scholar
  11. 11.
    Kurzynski, M., Zolnierek, A.: Rough sets and fuzzy sets theory applied to the sequential classification - algorithms and applications. Polish Journal of Environmental Studies 17(2B), 68–77 (2008)Google Scholar
  12. 12.
    Majak, M., Zolnierek, A.: Rough sets approach to the problems of classification. In: Proceedings of International Conference MOSIS X., pp. 109–114 (2010)Google Scholar
  13. 13.
    Zolnierek, A., Majak, M.: Rough sets approach to the classification task with modification of decision rules. In: Proceedings of the 11th WSEAS International Conference on Systems Theory and Scientific Computation, pp. 53–56 (2011)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  • Andrzej Zolnierek
    • 1
  • Marcin Majak
    • 1
  1. 1.Department of Systems and Computer Networks, Faculty of ElectronicsWroclaw University of TechnologyWroclawPoland

Personalised recommendations