Chapter 2: Introducing Hyperparameter Tuning
Every machine learning (ML) project should have a clear goal and success metrics. The success metrics can be in the form of business and/or technical metrics. Evaluating business metrics is hard, and often, they can only be evaluated after the ML model is in production. On the other hand, evaluating technical metrics is more straightforward and can be done during the development phase. We, as ML developers, want to achieve the best technical metrics that we can get since this is something that we can optimize.
In this chapter, we'll learn one out of several ways to optimize the chosen technical metrics, called hyperparameter tuning. We will start this chapter by understanding what hyperparameter tuning is, along with its goal. Then, we'll discuss the difference between a hyperparameter and a parameter. We'll also learn the concept of hyperparameter space and possible distributions of hyperparameter values that you may find...