SemTra: A semi-supervised approach to traffic flow labeling with minimal human effort
作者：
Highlights：
• Present a multi-view approach that creates multiple representations for traffic data.
• Incorporate different distance metrics in clustering to reveal the underlying structure.
• Explore the concept of evidence accumulation clustering.
• Propose mapping process based on the internal-structure and the probability function.
• Propose local and Global self-training approaches for predicting actual class label.
摘要
•Present a multi-view approach that creates multiple representations for traffic data.•Incorporate different distance metrics in clustering to reveal the underlying structure.•Explore the concept of evidence accumulation clustering.•Propose mapping process based on the internal-structure and the probability function.•Propose local and Global self-training approaches for predicting actual class label.
论文关键词：Internet traffic classification,Semi-supervised learning,Multiview
论文评审过程：Received 29 December 2016, Revised 31 January 2019, Accepted 4 February 2019, Available online 11 February 2019, Version of Record 18 February 2019.
论文官网地址：https://doi.org/10.1016/j.patcog.2019.02.001
