Network Anomaly Detection Based on Co-Training and Data Fusion
HUANG Yi-xiang;Software School of Tong Ji University;
With the development of network technology, network security has arised as one of the most serious problems. Classification methods Based on network behavior features always have good performance, however, in face of the multi-domain information of network traffic data, it's important to fuse them together efficiently. Besides, it's hard to train a proper model due to the lack of labeled abnormal IP addresses. This paper introduces an anomaly detection model Based on Co-Training and data fusion method. Through experiments on real data, the results prove that this method well solve the lack of ground truth under the premise of ensuring detecting accuracy.
【CateGory Index】： TP393.08