Abstract

In this era of Internet every electronic device being manufactured comes with built-ins to connect itself with the Internet. Such electronic gadgets create a massive amount of traffic flows every second. Managing this immense Internet traffic has become an exhausting job. Classifying these massive flows of data and interpreting them has attained considerable importance in the sight of researchers and Internet Service Providers. Many methods have been presented by various researchers to address this classification issue. This paper presents a Hybrid Classifier algorithm, which is a combination of the well-known machine learning types; the supervised learning and unsupervised learning to solve this classification issue. Accurate and Efficient recognition of flows is the key to manage flows in real-time. This algorithm classifies the traffic flows into real time flows and non-real time flows. It uses the built decision support model for classifying the flows based on the target class. To further validate the classification, it applies the k-means model. A significant improvement in the classification accuracy has been obtained.

Authors

P Shanmugaraja¹, K Chokkanathan², K Thangaraj³, P Ilanchezhian⁴
Sona College of Technology, India^1,3,4, Madanapalle Institute of Technology and Science, India²

Keywords

Voltage Stability, Voltage Collapse, Voltage Stability Indices, Radial Distribution Network