TY - JOUR
T1 - Design of adaptive fuzzy model for classification problem
AU - Li, Tzuu Hseng S.
AU - Guo, Nai Ren
AU - Kuo, Chao Lin
N1 - Funding Information:
Financial support from the National Science Council of the Republic of China under grants NSC92- 2212 -E006-063 and NSC92-2213 -E006-071 is gratefully acknowledged.
PY - 2005/4
Y1 - 2005/4
N2 - The main theme of this paper is to set up an adaptive fuzzy model for a new classification problem. At first, we propose a fuzzy classification model that can automatically generate the fuzzy IF-THEN rules by the features of the training database. The consequent part of the fuzzy IF-THEN rule consists of the confident value of the rule and which class the datum should belong to. Then a novel adaptive modification algorithm (AMA) is developed to tune the confident value of the fuzzy classification model. The proposed model comprises three modules, generation of the fuzzy IF-THEN rules, determination of the classification unit, and setup of the AMA. Computer simulations on the well known Wine and Iris databases have tested the performance. Simulations demonstrate that the proposed method can provide sufficiently high classification rate in comparison with other fuzzy classification models.
AB - The main theme of this paper is to set up an adaptive fuzzy model for a new classification problem. At first, we propose a fuzzy classification model that can automatically generate the fuzzy IF-THEN rules by the features of the training database. The consequent part of the fuzzy IF-THEN rule consists of the confident value of the rule and which class the datum should belong to. Then a novel adaptive modification algorithm (AMA) is developed to tune the confident value of the fuzzy classification model. The proposed model comprises three modules, generation of the fuzzy IF-THEN rules, determination of the classification unit, and setup of the AMA. Computer simulations on the well known Wine and Iris databases have tested the performance. Simulations demonstrate that the proposed method can provide sufficiently high classification rate in comparison with other fuzzy classification models.
UR - http://www.scopus.com/inward/record.url?scp=14844316699&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=14844316699&partnerID=8YFLogxK
U2 - 10.1016/j.engappai.2004.09.011
DO - 10.1016/j.engappai.2004.09.011
M3 - Article
AN - SCOPUS:14844316699
SN - 0952-1976
VL - 18
SP - 297
EP - 306
JO - Engineering Applications of Artificial Intelligence
JF - Engineering Applications of Artificial Intelligence
IS - 3
ER -