Temporal Attention-Guided Sequence Learning for Zero-Shot Generalization in Datacenter Network Failures
Abstract
Your research introduces an Attention-guided LSTM model to detect early warning signals in datacenter telemetry streams and predict failures before outages occur. The model is evaluated on optical network telemetry and BGP routing datasets and demonstrates improved F1-scores compared to classical machine learning baselines.
Fields of Study & Subjects
Software System Performance and Reliability
Network Security and Intrusion Detection
Software-Defined Networks and 5G
Publication details
- Journal / source
- Figshare
- Publication date
- 1 January 2026
- Language
- EN
- Document type
- Preprint