AI-Powered Business Intelligence

Chapter 3. Machine Learning Fundamentals

This chapter contains everything you need to know about machine learning—at least for this book. And it’s a great primer for the rest of your learning. The following sections should give you enough knowledge to follow along with the use cases in this book and help you build your own first prototypes. We’ll cover supervised machine learning, popular ML algorithms, and key terms, and you will learn how to evaluate ML models.

If you’re already familiar with these topics, feel free to consider this chapter a refresher. Let’s get started with supervised machine learning!

The Supervised Machine Learning Process

Let’s consider a simple example: imagine you want to sell your house and are wondering how to come up with a listing price. To get a realistic price, you would most likely look at other similar houses and the prices they were sold for. To come up with a good estimate, you would probably also compare your house to other houses in terms of some key features, such as overall size, bedrooms, location, and age. Without knowing it, you would have just acted like a supervised machine learning system.

Supervised machine learning is a process of training an ML model based on historical data when the ground truth is known. For example, if you want to estimate (or predict) the real estate price as in our example, a supervised learning algorithm would look at historical house prices (the label) and other information that describes the houses (the ...