Outliers Machine Learning
noun
Definition: Data points that differ substantially from the majority of observations in a dataset and may reflect rare events, variability, or data-quality issues [NIST/SEMATECH e-Handbook].
Example in context: “Figure 2 shows that all five manually added outliers have low changing scores regardless of using GDK or IDK.” [Cao et al. 2023]
Synonym: anomalies (context-dependent)
Related terms: outlier detection, anomaly detection, extreme values