From Data to Decision: A Multi-Stage Framework for Class Imbalance Mitigation in Optical Network Failure Analysis
This work addresses failure management for optical networks, but it is incremental as it compares existing techniques rather than introducing a new method.
The paper tackled class imbalance in optical network failure analysis by comparing pre-, in-, and post-processing methods, finding that post-processing improved F1 scores by up to 15.3% for detection and GenAI methods boosted performance by up to 24.2% for identification.
Machine learning-based failure management in optical networks has gained significant attention in recent years. However, severe class imbalance, where normal instances vastly outnumber failure cases, remains a considerable challenge. While pre- and in-processing techniques have been widely studied, post-processing methods are largely unexplored. In this work, we present a direct comparison of pre-, in-, and post-processing approaches for class imbalance mitigation in failure detection and identification using an experimental dataset. For failure detection, post-processing methods-particularly Threshold Adjustment-achieve the highest F1 score improvement (up to 15.3%), while Random Under-Sampling provides the fastest inference. In failure identification, GenAI methods deliver the most substantial performance gains (up to 24.2%), whereas post-processing shows limited impact in multi-class settings. When class overlap is present and latency is critical, over-sampling methods such as the SMOTE are most effective; without latency constraints, Meta-Learning yields the best results. In low-overlap scenarios, Generative AI approaches provide the highest performance with minimal inference time.