Silhouettes: a graphical aid to the interpretation and validation of cluster analysis
Journal of Computational and Applied Mathematics
Early application identification
CoNEXT '06 Proceedings of the 2006 ACM CoNEXT conference
Network Protocol Verification by a Classifier Selection Ensemble
MCS '09 Proceedings of the 8th International Workshop on Multiple Classifier Systems
Data clustering: 50 years beyond K-means
Pattern Recognition Letters
Hi-index | 0.00 |
In computer networks, the ability to correctly classify and control traffic flows is essential in order to manage network resources. A number of works have focused on the identification of flow attributes, or discriminators, able to distinguish the underlying application protocol of a flow at an early stage of it's existence. In this study k-means is investigated for identifying distinct application protocols present within flow data sets generated using a select number of discriminators. The clusters identified were used in a supervised training process that correctly identified protocols with an almost perfect (99% percent) success rate.