Journal of Intelligence and Information Systems (지능정보연구)

Korea Intelligent Information System Society (한국지능정보시스템학회)
권호 표지

Evaluation of Interpretability for Generated Rules from ANFIS

ANFIS에서 생성된 규칙의 해석용이성 평가

  • 송희석 (한남대학교 경영정보학과) ;
  • 김재경 (경희대학교 경영대학 & 경영연구원)
  • Received : 2009.11.17
  • Accepted : 2009.12.09
  • Published : 2009.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

Fuzzy neural network is an integrated model of artificial neural network and fuzzy system and it has been successfully applied in control and forecasting area. Recently ANFIS(Adaptive Network-based Fuzzy Inference System) has been noticed widely among various fuzzy neural network models because of outstanding performance of control and forecasting accuracy. ANFIS has capability to refine its fuzzy rules interactively with human expert. In particular, when we use initial rule structure for machine learning which is generated from human expert, it is highly probable to reach global optimum solution as well as shorten time to convergence. We propose metrics to evaluate interpretability of generated rules as a means of acquiring domain knowledge and compare level of interpretability of ANFIS fuzzy rules to those of C5.0 classification rules. The proposed metrics also can be used to evaluate capability of rule generation for the various machine learning methods.

퍼지신경망 모형은 인공신경망의 네트워크 구조 표현방법 및 학습알고리듬과 퍼지시스템의 추론방법을 통합한 모형으로 제어 및 예측분야에 성공적으로 적용되고 있다. 본 연구에서는 퍼지신경망 모형 중 우수한 예측정확도로 인해 최근 각광받고 있는 ANFIS (Adaptive Network-based Fuzzy Inference System) 모형에서 생성된 퍼지규칙의 해석용이성을 평가하였다. ANFIS모형은 인간 전문가와 상호작용하면서 규칙을 정제해 나갈 수 있다. 특히 인간전문가의 사전지식을 이용하여 초기 퍼지규칙을 만들고 난 후 모형을 학습하면 최적에 수렴하는 시간을 단축할 뿐 아니라, 전역 최적치 도달가능성이 높아진다고 보고되고 있다. 이러한 관점에서 볼 때 규칙의 해석용이성은 인간 전문가와의 상호작용을 위해 매우 중요한 이슈가 될 수 있다. 본 연구에서는 ANFIS모형과 의사결정나무 모형에서 생성된 규칙을 해석용이성 관점에서 비교하기 위한 측도를 제안하고 각 규칙들을 비교하였다. 본 연구에서 제안된 해석용이성 측도들은 규칙을 생성하는 다양한 기계학습 모형의 규칙생성 능력을 평가하는 기준으로도 활용될 수 있을 것이다.

Keywords

References

  1. Asuncion, A. and D. J. Newman, UCI Machine Learning Repository [http://www.ics.uci.edu/~mlearn/MLRepository.html]. Irvine, CA : University of California, School of Information and Computer Science, 2007.
  2. Babuska, R. and H. Verbruggen, "Neuro-fuzzy methods for nonlinear system identification", Annual Reviews in Control, Vol.27(2003), 73-85. https://doi.org/10.1016/S1367-5788(03)00009-9
  3. Bersini, H. and G. Bontempi, "Now comes the time to defuzzify neuro-fuzzy models", Fuzzy Sets and Systems, Vol.90(1997), 161-169. https://doi.org/10.1016/S0165-0114(97)00082-1
  4. Bodenhofer, U. and P. Bauer, A formal model of interpretability of linguistic variables : Trade-off between Accuracy and Interpretability in Fuzzy Rule-Based Modelling, Studies in Fuzziness and Soft Computing, Physica, Heidelberg, 2002.
  5. Chen, M.-Y. and D. A. Linkens, "Rule-base self-generation and simplification for data-driven fuzzy models", Fuzzy Sets and Systems, Vol.142, No.2(2004), 243-265. https://doi.org/10.1016/S0165-0114(03)00160-X
  6. Chiu, S., "Fuzzy Model Identification Based on Cluster Estimation", Journal of Intelligent and Fuzzy Systems, Vol.2, No.3(1994).
  7. Cordon, O. and F. Herrera, "A proposal for improving the accuracy of linguistic modeling", IEEE Trans. Fuzzy Systems Vol.8, No.3(2000), 335-344. https://doi.org/10.1109/91.855921
  8. Efendigil, T., S. Onut and C. Kahraman, "A decision support system for demand forecasting with artificial neural networks and neuro-fuzzy models : A comparative analysis", Expert Systems with Applications, Vol.36 (2009), 6697-707. https://doi.org/10.1016/j.eswa.2008.08.058
  9. Han M., Y. Sun and Y. Fan, "An improved fuzzy neural network based on T. S model", Expert Systems with Applications, Vol.34 (2008), 2905-2920. https://doi.org/10.1016/j.eswa.2007.05.020
  10. Jang, J.-S. R., "ANFIS: Adaptive-Network-based Fuzzy Inference Systems", IEEE Transactions on Systems, Man, and Cybernetics, Vol.23, No.3(1993), 665-685. https://doi.org/10.1109/21.256541
  11. Jang, J.-S. Roger and C.-T. Sun, "Neuro-Fuzzy Modeling and Control", The Proceedings of the IEEE, Vol.83, No.3(1995), 378-406. https://doi.org/10.1109/5.364486
  12. Jin, Y. W., V. Seelen and B. Sendhoff, "An approach to rule-based knowledge extraction", Proceedings of IEEE Conference on Fuzzy Systems, 1998, 1188-1193.
  13. Jin, Y. W., V. Seelen and B. Sendhoff, "On generating flexible, complete, consistens and compact (FC3) fuzzy rules from data using evolution strategies", IEEE Transactions on Systems, Man, and Cybernetics, Vol.29, No.4(1999), 829-845. https://doi.org/10.1109/3477.809036
  14. Jin, Y., "Fuzzy modeling of high-dimensional systems : complexity reduction and interpretability improvement", IEEE Transactions on Fuzzy Systems, Vol.8, No.2(2000), 212-221. https://doi.org/10.1109/91.842154
  15. Matlab, Fuzzy logic toolbox 2 user's guide. The Math Works Inc, 2009.
  16. Mikut, R., J. Jakel, and L. Groll, "Interpretability issues in data-based learning of fuzzy systems", Fuzzy Sets and Systems, Vol.150(2005), 179-197. https://doi.org/10.1016/j.fss.2004.06.006
  17. Nauck, D. and R. Kruse, "Neuro-fuzzy systems for function approximation", Fuzzy Sets and Systems, Vol.101(1999), 261-271. https://doi.org/10.1016/S0165-0114(98)00169-9
  18. Song, H. S. and J. K. Kim, "Design and Evaluation of ANFIS-based Classification Model", Journal of Inteligence and Information Systems, Vol.15, No.3(2009), 151-165.
  19. Takagi, T. and M. Sugeno, "Derivation of fuzzy control rules from human operator's control actions", Proceedings of the IFAC symposium on fuzzy information, knowledge representation and decision analysis, 1983, 55-60.
  20. Valente J., "Semantic constraints for membership function optimization", IEEE Trans, Systems Man Cybernetics.Part A : Systems and Humans Vol.29, No.1(1999), 128-138. https://doi.org/10.1109/3468.736369