detecting-data-anomalies

1# Detecting Data Anomalies
2 
3## Positioning
4 
5Treat this skill as an explicit/manual helper.
6In governed ML routing, anomaly-detection ownership normally belongs to `scikit-learn`.
7 
8## When to Use
9 
10Use this skill when:
11- Reviewing outlier transactions, fraud candidates, sensor spikes, or rare failures
12- Comparing isolation forest, one-class SVM, LOF, or threshold-based anomaly workflows
13- Turning suspicious records into a shortlist for human inspection
14 
15## Not For / Boundaries
16 
17- Null/duplicate/schema/range validation: use `exploratory-data-analysis`
18- Full model training or end-to-end pipeline ownership: use `scikit-learn` or `ml-pipeline-workflow`
19- Publication-grade figure production: use `scientific-visualization`
20 
21## Typical Outputs
22 
23- Candidate anomaly-detection methods and thresholds
24- A review checklist for false positives and false negatives
25- Suggested tables or plots for the suspicious subset
26 
27## Related Skills
28 
29- `scikit-learn` as the governed routed owner for classical anomaly-detection workflows
30- `creating-data-visualizations` after anomalies are identified