Introduction

In the previous chapters, we learned how to analyze and prepare a dataset in order to increase its level of quality. In this chapter, we will introduce you to another interesting topic: creating new features, also known as feature engineering. You already saw some of these concepts in Chapter 3, Binary Classification, but we will dive a bit deeper into this chapter.

The objective of feature engineering is to provide more information for the analysis you are performing on or the machine learning algorithms you will train on. Adding more information will help you to achieve better and more accurate results.

New features can come from internal data sources such as another table from databases or from different systems. For instance, you may want to link data from the CRM tool used in your company to the data from a marketing tool. The added features can also come from external sources such as open-source data or shared data from partners or providers. For example,...