A non-rules based approach to anomaly detection, leveraging the proven success of the NetHealth project. Enables ‘item level’ anomaly detection.
- Analyzes all elements of each data set it is applied to
- Anomaly detection based on previously observed behaviour
- All variances monitored in all files
- Low maintenance – no need to add/remove checks
- Analysis runs as close to data receipt as possible, giving the maximum amount of time for investigation and remediation of an issue before data release
The current Nethealth #2 Project is the second Innovation Partnership project and develops Machine Learning (ML) for Risk assessment.
Innovation Partnership 2 evolves Nethealth to ‘Pandora’ where Incoming Retailer Data allows feeds of QC Pre-Emptive Anomaly Detection for Outlet and Retailer Data.
“Pandora” for Retailers data deploys proprietary analytical algorithms, developed for network traffic anomaly detection, to retailer feeds with both “Nethealth” & “Pandora” machine learning anomaly detection & visualization feeds.
Xetec