Multivariate Outliers: Standardised Distance (Mahalanobis-Lite)
This lesson is part of the course
In this lesson
You standardise three transaction features, combine them into one distance score and find the payment that looks normal on every single feature but odd as a whole.
Buy the course once for $9.99 to unlock every lesson and lab, the final exam and the certificate – lifetime access, no subscription.
Lessons in this course
- Z-Score Flags: How Far from Normal Is a Transaction? Locked
- Robust Z-Scores with the Median & MAD Locked
- IQR Fences: Box-Plot Outliers in Server Latency Locked
- Rolling Mean ± k·σ Bands for a Live Metric (with a Chart) Locked
- Seasonal Baselines: Same Hour, Last Weeks Locked
- Multivariate Outliers: Standardised Distance (Mahalanobis-Lite) Locked
- Labelled Anomalies: Precision & Recall at Each Threshold (with a Chart) Locked
- Alert Fatigue & Cost: Choosing the Threshold Locked