Research in Clustering Algorithm for Diseases Analysis

• Published in: Journal of networks Vol. 8; no. 7; p. 1632
• Main Authors: XIA, Kaijian, WU, Yue, REN, Xiaogang, JIN, Yong
• Format: Journal Article
• Language: English
• Published: Oulu Academy Publisher 01.07.2013
• Subjects: Algorithms; Clustering; Clusters; Density; Diseases; Fuzzy; Fuzzy logic; Optimization
• ISSN: 1796-2056; 1796-2056
• DOI: 10.4304/jnw.8.7.1632-1639

The clustering algorithm plays a very important role in the applications of medical analysis, it can effective analysis the log of disease. It can accurately analyze the characteristics of various diseases, thus providing accurate basis for the doctor's diagnosis. In this paper, we will analysis the cluster algorithm-Fuzzy C-Means Algorithm (FCM). The traditional FCM is liable to trap into the problem of local optimum. We propose an improved algorithm of FCM based on the smooth technology. It will consider the sample points in different positions have different effects on cluster and cluster number has a great influence on the clustering results, so the new algorithm combines the point density and the method of determining the optimal number of clusters, finally use the effective evaluation function to evaluate the effective of the algorithm. Finally, we will use the case of Parkinson's disease to do the experimental verification, the results showed that the new clustering algorithms have better clustering effect, and it can more accurate analysis the characteristics of the disease, and it is using in some applications. Index Terms-Cluster Algorithm, Fuzzy C-Means Algorithm, Smooth Technology, Point density, Disease Analysis