What is a Point Anomaly?

Updated May 5, 2026

Short answer

A single data instance that is anomalous relative to the rest of the data[cite: 1].

Deep explanation

This is the simplest form of anomaly. If a single data point is far removed from the global distribution, it is a point anomaly[cite: 1].

Real-world example

A $10,000 purchase on a credit card that usually spends $20[cite: 1].

Common mistakes

  • Confusing it with contextual anomalies which require extra variables[cite: 1].

Follow-up questions

  • Can point anomalies be detected by Z-score?

More Anomaly Detection interview questions

View all →