Neuro-fuzzy-genetic Architecture for Data Mining
The main objective of this research is to develop a neuro-fuzzy-genetic architecture for data mining which presents discovered patterns in understandable form. In the architecture, all input variables are first preprocessed and all continuous variables are fuzzified. Fuzzification of the continuous inputs with linguistic terms represented by membership functions creates better understanding for the users. Principal Component Analysis (PCA) is then applied to reduce the dimension of the fuzzified input variables in finding combinations of variables, or factors, that describe major trends in the data. The rule extraction module in the architecture is designed to extract explicit knowledge from the trained neural networks and represents it in the form of crisp and fuzzy If-Then rules. In the final stage a genetic algorithm is used as a rule-pruning module to get rid of the weak rules that are still in the rule bases, while the classification accuracy of the rules bases is significantly changed or even improved. Benchmarking the performance of the architecture with the standard C4.5 decision tree was also carried out. Real world application of the architecture is demonstrated with the meningoencephalitis diagnosis dataset from the JSAI KDD Challenge 2001.
K. Hemsathapat et al., "Neuro-fuzzy-genetic Architecture for Data Mining," Intelligent Engineering Systems Through Artificial Neural Networks, American Society of Mechanical Engineers (ASME), Jan 2001.
Engineering Management and Systems Engineering
Keywords and Phrases
Data Mining; Fuzzy Sets; Genetic Algorithms; Neuro-Fuzzy-Genetic Data Mining Architecture
Article - Conference proceedings
© 2001 American Society of Mechanical Engineers (ASME), All rights reserved.