Abstract
In this work, we review one of the unsupervised clustering methods to provide low-rank parametric embeddings of network activity patterns. Since optical parameters exhibit spatial and temporal locality, this embedding based on time-series correlation coefficients of optical parameters. To this end, we use hierarchical clustering to create dendrogram representation for different network activity patterns or repeating trends.
© 2019 The Author(s)
PDF ArticleMore Like This
M. M. Hasan and Jason P. Jue
OThQ2 Optical Fiber Communication Conference (OFC) 2008
Elias Giacoumidis, Amir Matin, Jinlong Wei, Nick J. Doran, and Xu Wang
Su4C.4 Asia Communications and Photonics Conference (ACP) 2017
Shoudong Liu, Jie Zhang, Yajie Li, Yongli Zhao, Kai Wang, Xiaokun Yang, and Yi Sun
T4B.6 Asia Communications and Photonics Conference (ACP) 2019