Reproduce “Practical Anomaly Detection over Multivariate Monitoring Metrics for Online Services” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jinyang Liu, Tianyi Yang, Zhuangbin Chen, Yuxin Su, Cong Feng, Zengyin Yang, Michael R. Lyu
A framework called CMAnomaly detects anomalies in multivariate monitoring metrics for large-scale online services by employing a collaborative machine that models metric interactions with linear time complexity. Deployed in Huawei Cloud, it achieved superior F1-scores and up to 20x faster training times compared to baselines.
Key insights
- Explicitly modeling interactions among multivariate metrics, across both feature and temporal dimensions, is crucial for accurate anomaly detection in complex systems.
- Computational efficiency in modeling these interactions can be achieved by reformulating them using factorization machine principles, reducing complexity from quadratic to linear time.
- A forecasting-based approach with multi-layer perceptrons, combined with efficient interaction modeling, offers a practical balance of accuracy and speed for industrial-scale deployment.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.