作者: Lili Yang , Jie Wang , Ping Zhong
DOI: 10.1007/978-3-319-11197-1_47
关键词:
摘要: Signature-based intrusion detection system is currently used widely, but it dependent on high quality and complete attack signature database. Despite a great number of automatic feature extraction has been proposed, however, with the progress technology, generation research still an open problem. This paper presents novel combining supervised unsupervised learning for based transport layer network statistics feature, outputs sets in feedback way. Finally we demonstrate effectiveness model by using data from laboratory Darpa2000 datasets.