点击次数：

发布时间：2019-03-09

论文类型：期刊论文

发表时间：2011-04-01

发表刊物：NONLINEAR DYNAMICS

收录刊物：EI、SCIE、Scopus

卷号：64

期号：1-2

页面范围：127-130

ISSN号：0924-090X

关键字：One class support vector machine; Anomaly detection; Local density degree

摘要：Anomalies are patterns in data that do not conform to a well-defined notion of normal behavior. One-class Support Vector Machines calculate a hyperplane in the feature space to distinguish anomalies, however, it may not identify the ideal hyperplane especially when the support vectors do not have the overall characteristics of the target data. So, we propose a new local density OCSVM by incorporating distance measurements based on local density degree to reflect the distribution of a given data set. Experimental results on UCI data sets show that the proposed method can achieve better performance than other one class learning schemes.