arXiv:2501.18196v3 Announce Type: replace Abstract: Unsupervised anomaly detection of multivariate time series is a challenging task, given the requirements of deriving a compact detection criterion without accessing the anomaly points. The existing methods are mainly based on reconstruction error or association divergence, which are both confined to isolated subsequences with limited horizons, hardly promising unified series-level criterion. In this paper, we propose the Global Dictionary-enhanced Transformer (GDformer) with a renovated dictionary-based cross attention mechanism to cultivate
The continuous growth of multivariate time series data in various applications necessitates more robust and comprehensive anomaly detection methods that move beyond fragmented analysis.
Improved anomaly detection in complex time series data can enhance the reliability and efficiency of AI systems, impacting critical infrastructures and decision-making processes across sectors.
The proposed GDformer offers a more unified and comprehensive approach to anomaly detection by considering global rather than isolated patterns, potentially leading to more accurate and contextually aware AI performance.
- · AI developers and researchers
- · Industries relying on sensor data (e.g., manufacturing, finance, health)
- · Cloud infrastructure providers
- · Legacy anomaly detection systems
- · Systems highly prone to false positives/negatives
More reliable and less error-prone AI systems leveraging time series data for critical applications.
Increased adoption of AI in sectors requiring high precision anomaly detection, leading to greater automation and operational efficiency.
The development of new AI applications built upon robust anomaly detection capabilities, potentially decentralizing monitoring and control in complex systems.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG