Summary

In this chapter, we have discussed anomaly detection. We began with univariate anomaly detection, including non-parametric and parametric approaches. We discussed performing data transformations to obtain better classifications of outliers. We then discussed multivariate anomaly detection using Mahalanobis distances. We completed more advanced exercises to classify anomalies related to seasonally varying data. We discussed collective and contextual anomalies, and concluded the chapter with a discussion of how to use kernel density estimation in anomaly detection.