The document discusses anomaly detection in data mining. It defines anomalies as data points that are considerably different than most other data points. It describes challenges in anomaly detection like determining how many outliers exist and finding outliers hidden among large amounts of normal data. It also discusses different approaches to anomaly detection, including graphical, statistical, distance-based, and clustering-based methods. Finally, it notes the importance of considering base rates when evaluating anomaly detection systems to avoid the base rate fallacy.