Outlier detection is the identification of rare items, events, or observations which raise suspicions by differing significantly from the majority of the data.
Crucial for robust statistical modeling.
Used in fraud detection, system health monitoring, and data cleaning.