Abstract
In multi-tiered anomaly detection systems, a challenge may exist in balancing the operational cost of accurate analytical models with the accuracy of less expensive rule engines, where manual rule creation can be slow. A described technology can address this by, for example, monitoring a multi-stage processing cascade and employing multi-stage failure learning. A system can identify instances where a low-cost analysis tier does not detect an anomaly that a higher-cost tier subsequently identifies. The system may then extract the successful detection pattern and use a generative model to formulate a new, low-cost rule that codifies this logic. This process can enable the deployment of learned intelligence to earlier processing tiers, which may reduce reliance on computationally expensive analysis for recurring patterns, help optimize operational costs, and improve adaptive capabilities.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Mitra, Subhadip, "Automated Rule Generation for Tiered Systems Using Multi-Stage Failure Learning", Technical Disclosure Commons, (January 12, 2026)
https://www.tdcommons.org/dpubs_series/9180