Advertisement

Balancing Interpretability against Accuracy in Fuzzy Modeling by Means of ACO

  • Pablo Carmona
  • Juan Luis Castro
  • José Luis Herrero
Part of the Communications in Computer and Information Science book series (CCIS, volume 297)

Abstract

Ant colony optimization (ACO) techniques have been revealed as an effective way to improve the interpretability of fuzzy models by reformulating an initial model [1,2]. However, despite this reformulation preserves the initial fuzzy rules, new rules can be added in the search for an interpretability enhancement. Thus, differences between the inferences of the initial and final models can arise due to the interaction among the initial rules and their adjacent new rules. This can lead to changes in the accuracy of the initial model. In order to keep the accuracy of the initial fuzzy model, this work proposes to include the difference between the outputs of the initial and final models as an additional criterion to evaluate the optimality of the reformulated model. This will allow to balance the interpretability against the accuracy within the optimization algorithm, even making the interpretability improvement conditional on a strict preservation of the initial accuracy.

Keywords

fuzzy modeling ant colony optimization interpretability vs. accuracy trade-off distortion measure 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Carmona, P., Castro, J.L.: Using ant colony optimization for learning maximal structure fuzzy rules. In: Proc. FUZZ-IEEE 2005, Reno, Nevada, pp. 702–707 (2005)Google Scholar
  2. 2.
    Carmona, P., Castro, J.L.: Mecanismos de mejora para la identificación de modelos difusos interpretables mediante algoritmos de hormigas. In: Actas del II Simposio sobre Lógica Fuzzy y Soft Computing, Zaragoza, Spain, pp. 30–44 (2007)Google Scholar
  3. 3.
    Alonso, J.M., Magdalena, L.: Special issue on interpretable fuzzy systems. Inf. Sci. 181(20), 4331–4339 (2011)MathSciNetCrossRefGoogle Scholar
  4. 4.
    Casillas, J., Cordón, O., Herrera, F., Magdalena, L.: Interpretability improvements to find the balance interpretability-accuracy in fuzzy modeling: An overview. In: Casillas, J., Cordón, O., Herrera, F., Magdalena, L. (eds.) Accuracy Improvements in Linguistic Fuzzy Modeling. STUDFUZZ, vol. 129, pp. 3–22. Springer, Heidelberg (2003)Google Scholar
  5. 5.
    Alonso, J.M., Guillaume, S., Magdalena, L.: A hierarchical fuzzy system for assessing interpretability of linguistic knowledge bases in classification problems. In: Proc. IPMU 2006, Paris, France, pp. 348–355 (2006)Google Scholar
  6. 6.
    Castro, J.L., Castro-Schez, J.J., Zurita, J.M.: Learning maximal structure rules in fuzzy logic for knowledge acquisition in expert systems. Fuzzy Sets Syst. 101, 331–342 (1999)MathSciNetzbMATHCrossRefGoogle Scholar
  7. 7.
    Dorigo, M., Stützle, T.: Ant Colony Optimization. MIT Press (2004)Google Scholar
  8. 8.
    Kaufmann, A., Gupta, M.M.: Introduction to Fuzzy Arithmetic, Theory and Applications. Van Nostrand Reinhold, New York (1985)zbMATHGoogle Scholar
  9. 9.
    Dorigo, M., Gambardella, L.M.: Ant colony system: A cooperative learning approach to the traveling salesman problem. IEEE Trans. Evol. Comput. 1(1), 53–66 (1997)CrossRefGoogle Scholar
  10. 10.
    Carmona, P., Castro, J.L.: Algoritmo de hormigas mejorado para la descripción de modelos difusos interpretables mediante reglas con excepciones. In: Actas ESTYLF 2008, Langreo-Mieres, Spain, pp. 185–190 (2008)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  • Pablo Carmona
    • 1
  • Juan Luis Castro
    • 2
  • José Luis Herrero
    • 1
  1. 1.Department of Computer and Telematics Systems EngineeringUniversity of ExtremaduraBadajozSpain
  2. 2.Department of Computer Science and Artificial IntelligenceUniversity of GranadaGranadaSpain

Personalised recommendations