Combining SOM and local minimum enclosing spheres for novelty detection

作     者：Hong-Jie Xing,Ming-Hu Ha,Xi-Zhao Wang College of Mathematics and Computer Science,Hebei University,Baoding,Hebei Province,China 

会议名称：《2009中国控制与决策会议》

会议日期：2009年

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 081104[工学-模式识别与智能系统] 08[工学] 0835[工学-软件工程] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：the National Natural Science Foundation of China(No.60773062) the Key Scientific and Technical Research of the Ministry of Education of China(No.206012) the China Postdoctoral Science Foundation(No.20080440820) the Scientific Research Project of Department of Education of Hebei Province(No.2008306) the Postdoctoral Science Foundation of Hebei University the Science Foundation of Hebei University(No.Y2008123) the KeyProject Foundation of Applied Fundamental Research of Hebei Province(08963522D) the Science and Technology Supporting Project of Scienceand Technology of Hebei Province(No.072135188) 

关 键 词：Self-Organizing Map Local Minimum Enclosing Spheres Novelty Detection 

摘      要：正In this paper,a novelty detection method based on self-organizing map（SOM） and local minimum enclosing spheres is *** are two phases in the proposed *** the first phase,the whole training set are split into disjointed Voronoi regions by *** the second phase,several local minimum enclosing spheres are constructed upon these Voronoi *** with its related works,the proposed method demonstrates better performances on one synthetic data set and two benchmark data sets.