An Improved Intrusion Detection Algorithm Based On DBSCAN
Jian Yang · Microcomputer Information · 2009
DBSCAN is an efficient analysis method of clustering algorithm in data mining. In this paper, a model of intrusion detection system is built, and an IDBC algorithm is applied in the designing of detection engine. IDBC algorithm improves the computing of distance of network connecting records, and clusters merges based on the result of DBSCAN. Experiment results prove that, comparing DBSCAN, IDBC reduces the false positive rates, and promote the performance of intrusion detection system.