Relative Network Entropy Based Clustering Algorithm for Intrusion Detection

Download Now
Provided by: International Journal of Network Security
Topic: Security
Format: PDF
Clustering, as a kind of data mining methods, with the characteristic of no supervising, quick modeling is widely used in intrusion detection. However, most of the traditional clustering algorithms use a single data point as a calculating unit and the drawback exists in time wasting to calculate one data point after another when clustering, meanwhile, a single local change of data will significantly affect the clustering results. This paper proposes a novel clustering algorithm named Entropy-Based DBSCAN (EB-DBSCAN), a data mining algorithm based on relative network entropy.
Download Now

Find By Topic