Data Mining: Concepts and Techniques, 3rd Edition

12.11 Bibliographic Notes

Hawkins [Haw80] defined outliers from a statistics angle. For surveys or tutorials on the subject of outlier and anomaly detection, see Chandola, Banerjee, and Kumar [CBK09]; Hodge and Austin [HA04]; Agyemang, Barker, and Alhajj [ABA06]; Markou and Singh [MS03a, MS03b]; Patcha and Park [PP07]; Beckman and Cook [BC83]; Ben-Gal [B-G05]; and Bakar, Mohemad, Ahmad, and Deris [BMAD06]. Song, Wu, Jermaine, et al. [SWJR-07] proposed the notion of conditional anomaly and contextual outlier detection.

Fujimaki, Yairi, and Machida [FYM05] presented an example of semi-supervised outlier detection using a set of labeled “normal” objects. For an example of semi-supervised outlier detection using labeled outliers, see Dasgupta and ...