期刊名称：International Journal of Computer Science Issues
印刷版ISSN：1694-0784
电子版ISSN：1694-0814
出版年度：2011
卷号：8
期号：5
出版社：IJCSI Press
摘要：A hybrid data mining model is proposed for finding an optimal number of different pathological types of any disease, and extracting the most significant features for each pathological type. This model is improved in order to reach the fewer subsets of features that have the most impact distinctive of each pathological type. This improvement is lead to the great importance in the decision making of the diagnosis process without confusion or ambiguity between the different variations of the diseases. This model and its optimization are based on fuzzy clustering, nearest neighbor classification, sequential backward search method, and averaging schema for features selection. Experiments have been conducted on three real medical datasets that have different diagnoses. The results show that the highest classification performance is obtained using our optimized model, and this is very promising compared to Navebayes, Linear and Polykernal Support Vector Machine (SVM), Artificial Neural Network (ANN), and Support Feature Machines (SFM) models.
关键词：Data Mining; Fuzzy Clustering; Nearest Neighbor Classification; Features Selection.
