The Study of Network Traffic Identification Based on Machine Learning Algorithm

Network traffic identification is one of the hot research fields for network management and network security; machine learning is an important method during the network traffic identification research.this paper describes the current situation and common methods of network traffic identification, at the same time this paper also states the currently popular Machine learning methods. We compared and evaluated the supervised and unsupervised classification and clustering algorithms, the experiment results show that feature selection algorithm has great effect on supervised machine learning and DBSCAN algorithm which belongs to unsupervised clustering algorithm has great potential in precision.