On Combining Neuro-Fuzzy Architectures with the Rough Set Theory to Solve Classification Problems with Incomplete Data
Issue No. 09 - September (2008 vol. 20)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/TKDE.2008.64
Robert Nowicki , Czestochowa University of Technology, Czestochowa
This paper presents a new approach to fuzzy classification in the case of missing features. The rough set theory is incorporated into neuro-fuzzy structures and the rough-neurofuzzy classifier is derived. The architecture of the classifier is determined by the MICOG (modified indexed center of gravity) defuzzification method. The structure of the classifier is presented in a general form which includes both the Mamdani approach and the logical approach - based on the genuine fuzzy aplications. A theorem, which allows to determine the structures of a roughneuro-fuzzy classifiers based on the MICOG defuzzification, is given and proven. Specific rough-neuro-fuzzy structures based on the Larsen rule, the Reichenbach and the Kleene-Dienes implications are given in details. In the experiments it is shown that the classifier with the Dubois-Prade fuzzy implication is characterized by the best performance in the case of missing features
Decision support, Rule-based processing, Uncertainty, "fuzzy", and probabilistic reasoning, Fuzzy set, Classifier design and evaluation
R. Nowicki, "On Combining Neuro-Fuzzy Architectures with the Rough Set Theory to Solve Classification Problems with Incomplete Data," in IEEE Transactions on Knowledge & Data Engineering, vol. 20, no. , pp. 1239-1253, 2008.