Semi-Supervised Machine Learning Algorithms for Classifying the Imbalanced Protocol Flows

摘要

The samples marked bottlenecks and imbalanced protocol flows restrict the development of the network traffic classification technology, to solve this problem a semi-supervised machine learning traffic identification method is presented. Employ K-means algorithm to partition a training datasets that consists of a few labeled flows combined with abundant unlabeled flows. Then, identify the unlabeled samples using the labeled samples in the cluster based on k Nearest Neighbor algorithm. The theoretical analysis and experimental results show that the algorithm can improve the recognition rate of minority flows in the case of the imbalanced protocol flows.