The Infuence Of Noisy Data On Skype Traffic Classification

DOI

10.2991/icaise.2013.52How to use a DOI?

Keywords

Component: Skype Traffic Classification, Neural Networks, C4.5, Noisy data

Abstract

Because of its popularity, encrypted traffic and proprietary design, there has been difficult to detect Skype from other P2P traffics. The research of Skype traffic identification focuses on collecting traffic flow feature and using machine learning method to identification. The key of machine learning method is datasets and flow feature selection. Since there is no publicly available datasets, noisy data can’t be avoided. In this paper, I compare two different machine learning classification techniques, C4.5 and Neural Networks. Results show that C4.5 is better than Neural Networks when noisy data percent is low and Neural Networks is steady when noisy data percent is high.

Copyright

© 2013, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Linhua Niu
AU  - Xiangzhan Yu
AU  - Zhimin Yin
PY  - 2013/08
DA  - 2013/08
TI  - The Infuence Of Noisy Data On Skype Traffic Classification
BT  - Proceedings of the 2013 The International Conference on Artificial Intelligence and Software Engineering (ICAISE 2013)
PB  - Atlantis Press
SP  - 242
EP  - 245
SN  - 1951-6851
UR  - https://doi.org/10.2991/icaise.2013.52
DO  - 10.2991/icaise.2013.52
ID  - Niu2013/08
ER  -