Chapter 6: Hyperparameter Tuning
In this chapter, we are going to start by looking at three different hyperparameter tuning algorithms—Hyperband, Bayesian optimization, and random search. These algorithms are implemented in the tf.keras API, which makes them relatively easy to understand. With this API, you now have access to simplified APIs for these complex and advanced algorithms that we will encounter in this chapter. We will learn how to implement these algorithms and use the best hyperparameters we can find to build and train an image classification model. We will also learn the details of its learning process in order to know which hyperparameters to search and optimize. We will start by getting and preparing the data, and then we'll apply our algorithm to it. Along the way, we will also try to understand key principles and the logic to implement user choices for these algorithms as user inputs, and we'll look at a template to submit tuning...