Clustering Based Network Intrusion Detection Using KDD Train 20 Percent

Provided by: International Journal of Advanced Research in Computer Science and Software Engineering (IJARCSSE)
Topic: Security
Format: PDF
In this Paper a clustering algorithm is proposed to work on network intrusion data. The algorithm is experimented with KDD Train 20 percent dataset and found satisfactory results. The authors perform clustering to group training data points into clusters, from which they select some clusters as normal and known-attack profile according to certain criterion. For those training data excluded from the profile, they use them to build a specific classifier. During the testing stage, they utilize influence based classification algorithm to classify network behaviors.

Find By Topic