Anomaly detection in customs area (EN)

Anomaly detection in customs area (EN)

16:45 - 17:15 | data.stage | Transport & Industry

The search for anomalous transactions was performed manually on historical data. A check for previously unknown anomalies and attribute constellations is hardly possible with manually defined rules. The aim of the project is to automatically detect anomalies and unwanted deviations using intelligent and learning procedures. The project therefore focused on advanced technologies (e.g. neural networks).

Michael Beckmann, Tax Manager & IT, Henkel AG & Co. KGaA