Machine learning-based selection of metrics for anomaly detection
US11748568B1 · kind B1 · utility
4Cited by
6References
20Claims
0Family size
Assignee
Inventors
Key dates
| Filing date | Aug 7, 2020 |
| Grant date | Sep 5, 2023 |
| Priority date | — |
| Expiry date | Dec 4, 2040 |
Classification
- Technology area (CPC H)Electricity
- CPC primaryH04L43/024
- WIPO fieldDigital communication
- WIPO sectorElectrical engineering
Abstract
A plurality of metrics records, including some records indicating metrics for which anomaly analysis has been performed, is obtained. Using a training data set which includes the metrics records, a machine learning model is trained to predict an anomaly analysis relevance score for an input record which indicates a metric name. Collection of a particular metric of an application is initiated based at least in part on an anomaly analysis relevance score obtained for the particular metric using a trained version of the model.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.