Chapter 7. Anomaly Detection
"Find the odd one out automatically."
Anomaly Detection is the art of finding the odd one out in a bunch of data automatically. As the name suggests, it is the science of finding the data that is anomalous compared to the other. However, there are several kinds of anomalies; sometimes the normal (non-anomalous) data can be line anomalous data. Anomaly detection is mostly an unsupervised learning problem because it is very difficult, if not impossible, to get a labeled training dataset that is anomalous. Sometimes, anomalies are referred to as "outliers."
