Abstract
Proximal events posing risks to network service were classified using Decision Trees on State of Polarization Multivariate Time Series data. Aggregate features of interests were individually evaluated to determine their significance, demonstrating that a combination of two aggregates sufficed to produced 98.8% event classification accuracy.
© 2022 The Author(s)
PDF ArticleMore Like This
Cen Wang, Hong Cao, Shenzhen Yang, Junyuan Guo, Hongxiang Guo, and Jian Wu
W1C.4 Optical Fiber Communication Conference (OFC) 2018
Yangmin Ding, Yue Tian, Sarper Ozharar, Zhuocheng Jiang, and Ting Wang
SM2C.7 Optical Sensors (Sensors) 2022
Ming-Fang Huang, Zhuocheng Jiang, Jian Fang, Shaobo Han, Ezra Ip, Sarper Ozharar, Yuheng Chen, Giovanni Milione, and Ting Wang
FM4D.1 Frontiers in Optics (FiO) 2022