The goal of this first part of the book is to answer the question, “What is an anomaly?” Throughout this part of the book, we will learn what an anomaly is, some of the common use cases for anomaly detection, and just how powerful is that anomaly detection engine between your ears. We will cover a variety of psychological, technical, and statistical concepts, and by the end of this first part, we will have enough knowledge of anomalies and anomaly detection to begin writing code.

In Chapter 1, we will learn about outliers and anomalies as well as use cases for anomaly detection and one way to classify different types of anomaly detection techniques. Then, in Chapter 2, we will dive into psychology, specifically the Gestalt school of psychology. We will review some of the research around how the human brain reacts to patterns, as well as good ways to share information with people. Chapter 3 moves from the human aspect to computers. Humans are innately great pattern matchers, but computers are not. What we can do is train machines to understand what is normal and what deviates from normal. This third chapter provides us some of the statistical techniques we can use to train a computer to detect outliers.