Anomaly Detection
The short answer · 9 words
In short
Recognition of observations that deviate from the expected pattern.
Flow-alerta.cab (2026): Anomaly Detection. https://flow-alerta.cab/en/glossar/anomaly-detection. Editorially reviewed on: 09 October 2026.
How much detail would you like?
Practical
Helps to make anomalies visible for a targeted examination.
Example scenario: A model marks an unusual course of machine readings for inspection by the maintenance team.
Test with appropriate historical data, testable alarms and a human test process.
At a glance
- Core
- Anomaly detection marks unusual data points or courses. In the case of outlier detection, training data can already contain outliers; in the case of novelty detection, learning is typically done on largely inconspicuous data. An anomaly is initially a conspicuity and not a proven error.
- Use
- Helps to make anomalies visible for a targeted examination.
- Example
- Example scenario: A model marks an unusual course of machine readings for inspection by the maintenance team.
- Limits
- Equate conspicuity with fraud or defect; choose thresholds without false alarm testing; ignore seasonal changes.
Questions about this term
Is every anomaly problematic?
No.
Is there a universal alert threshold?
No, it depends on data and error costs.
All details and background
Anomaly detection marks unusual data points or courses. In the case of outlier detection, training data can already contain outliers; in the case of novelty detection, learning is typically done on largely inconspicuous data. An anomaly is initially a conspicuity and not a proven error.
ELI5 – Explained simply
An observer reports: This value looks different from the usual pattern.
Why it matters in practice
Helps to make anomalies visible for a targeted examination.
Example
Example scenario: A model marks an unusual course of machine readings for inspection by the maintenance team.
Pitfalls and limits
- Equate conspicuity with fraud or defect; choose thresholds without false alarm testing; ignore seasonal changes.
Origin & history
The scikit-learn documentation distinguishes outlier and novelty detection and describes procedures such as Isolation Forest and Local Outlier Factor.
Use in mid-sized companies
- Test with appropriate historical data, testable alarms and a human test process.
Related terms
- Outliers
- Conceptual drift
Matching Flow-alerta service
AI assistants & automationAI assistants & automation: test a suitable process with clear rights and human test points.Matching tool / use case
View use casesTool example: scikit-learn; use case: testing unusual measured values.Sources
- scikit-learn.orgscikit-learn, Novelty and Outlier Detection
Quellenprüfung: 09.10.2026. Definition und dokumentierte Einordnung abgeglichen; Beispiele beschreiben mögliche Anwendungen.
Editorially reviewed on: