Anomaly Detection
Anomaly detection is the use of AI to identify data points, transactions or events that deviate significantly from the expected pattern. Rather than being told what a problem looks like, the model learns what "normal" is and flags what does not fit.
In logistics and finance it surfaces duplicate or fraudulent invoices, mis-declared weights, unusual costs, and shipments behaving abnormally — catching issues that rule-based checks miss, before they become losses.
Also known as
Outlier Detection
Related terms
Where this matters at WHIZTEC
Frequently asked
How is anomaly detection different from a rule?
A rule only catches what you defined in advance. Anomaly detection learns normal behaviour and flags anything unusual — including problems no one thought to write a rule for.