Research Article
Traffic Classification Based on Incremental Learning Method
@INPROCEEDINGS{10.1007/978-3-319-73317-3_40, author={Guanglu Sun and Shaobo Li and Teng Chen and Yangyang Su and Fei Lang}, title={Traffic Classification Based on Incremental Learning Method}, proceedings={Advanced Hybrid Information Processing. First International Conference, ADHIP 2017, Harbin, China, July 17--18, 2017, Proceedings}, proceedings_a={ADHIP}, year={2018}, month={2}, keywords={Traffic classification Incremental learning Support vector machine}, doi={10.1007/978-3-319-73317-3_40} }- Guanglu Sun
Shaobo Li
Teng Chen
Yangyang Su
Fei Lang
Year: 2018
Traffic Classification Based on Incremental Learning Method
ADHIP
Springer
DOI: 10.1007/978-3-319-73317-3_40
Abstract
Machine learning methods become more and more important in traffic classification, because they are able to explore statistical features to identify encrypted traffic and proprietary protocols. Among many machine learning methods, support vector machine is able to achieve state of the art performance in classifying TCP traffic. However, current support vector machine for traffic classification also shows two limitations: (i) unable to support continuously learning, and (ii) high requirements on both memory and CPU. In this paper, incremental Support Vector Machine method is applied to address these two issues. Experimental results show that the incremental Support Vector Machine method decreases the training time, while still sustains the high accuracy of traffic classification.
