Beginning Anomaly Detection Using

CHAPTER 2

Traditional Methods 

of Anomaly Detection

In this chapter, you will learn about traditional methods of anomaly detection. You 

will also learn how various statistical methods and machine learning algorithms work 

and how they can be used to detect anomalies and how you can implement anomaly 

detection using several algorithms.

In a nutshell, the following topics will be covered throughout this chapter:

•  A data science review

•  The three styles of anomaly detection

•  The isolation forest

•  One-class support vector machine (OC-SVM)

 Data Science Review

It is important to understand some basic data science concepts in order for you to 

evaluate how well your model performs and to compare its performance with other 

models.

First of all, the goal in anomaly detection is to determine whether or not a given 

point is an anomaly or not. Essentially, you are labeling a data point