ABSTRACT:

There is vital interest in the info mining and network management communities about the necessity to improve existing techniques for clustering multi-variate network traffic flow records therefore that we have a tendency to will quickly infer underlying traffic patterns. In this paper we have a tendency to investigate the use of clustering techniques to spot interesting traffic patterns from network traffic information in an efficient manner. We tend to develop a framework to house mixed type attributes as well as numerical, categorical and hierarchical attributes for a one-pass hierarchical clustering algorithm. We tend to demonstrate the improved accuracy and efficiency of our approach compared to previous work on clustering network traffic.