Applying Rough Sets Reduction Techniques to the Construction of a Fuzzy Rule Base for Case Based Reasoning
Early work on Case Based Reasoning reported in the literature shows the importance of soft computing techniques applied to different stages of the classical 4-step CBR life cycle. This paper proposes a reduction technique based on Rough Sets theory that is able to minimize the case base by analyzing the contribution of each feature. Inspired by the application of the minimum description length principle, the method uses the granularity of the original data to compute the relevance of each attribute. The rough feature weighting and selection method is applied as a pre-processing step previous to the generation of a fuzzy rule base that can be employed in the revision phase of a CBR system. Experiments using real oceanographic data show that the proposed reduction method maintains the accuracy of the employed fuzzy rules, while reducing the computational effort needed in its generation and increasing the explanatory strength of the fuzzy rules.
Unable to display preview. Download preview PDF.
- 1.Riesbeck, C.K., Schank, R.C.: Inside Case-Based Reasoning. Lawrence Erlbaum Associates, Hillsdale (1999)Google Scholar
- 2.Pal, S.K., Dilon, T.S., Yeung, D.S.: Soft Computing in Case Based Reasoning. Springer, London (2000)Google Scholar
- 3.Sankar, K.P., Simon, C.K.S.: Foundations of Soft Case-Based Reasoning. Wiley- Interscience, Hoboken (2003)Google Scholar
- 10.Díaz, F., Corchado, J.M.: A method based on the Rough Set theory and the MDL principle to select relevant features. In: Proc. of the X CAEPIA - V TTIA, vol. 1, pp. 101–104 (2003)Google Scholar
- 11.Rissanen, J.: Minimum description length principle. In: Kotz, S., Johnson, N.L. (eds.) Encyclopedia of Statistical Sciences, pp. 523–527. John Wiley and Sons, New York (1985)Google Scholar
- 14.Fdez-Riverola, F., Corchado, J.M.: An automated CBR Revision method based on a set of β-TSK Fuzzy models. In: Proc. of the X CAEPIA - V TTIA, vol. 1, pp. 395–404 (2003)Google Scholar
- 15.Fdez-Riverola, F.: Neuro-symbolic model for unsupervised forecasting of changing environments. Ph.D. diss., Dept. of Computer Science, Vigo University, Spain (2002)Google Scholar