Presentation Information

[ThG1-6]Prior Knowledge-Assisted Self-Supervised Learning for Soft Fault Diagnosis with Limited Fault Samples in Optical Networks

○Xue Xiao, Chunyu Zhang, Min Zhang, Qian Guo, Zhiming Sun, Danshi Wang (Beijing Univ. of Posts and Telecommunications)

Keywords:

Artificial intelligence and machine learning for optical transmission systems and subsystems

We proposed a prior knowledge-assisted self-supervised learning scheme for soft fault diagnosis in optical networks with limited fault samples, achieving the highest average F1-score of 0.9902 and FNR of 0.0891 through prior knowledge embedding.