Introduction
In the previous chapter, you were introduced to TensorFlow and Keras, along with an overview of their key features and applications and how they work in synergy. You learned how to implement a deep neural network with TensorFlow, addressing all major topics, that is, model creation, training, validation, and testing, using the most advanced machine learning frameworks available. In this chapter, we will use this knowledge to build models that are able to solve some classical reinforcement learning problems.
Reinforcement learning is a branch of machine learning that comes closest to the idea of artificial intelligence. The goal of training an artificial system to learn a given task, without any prior information, and only by means of experiences of an environment, represents the ambitious aim of replicating human learning. Applying deep learning techniques to the field has recently led to a great increase in performance, thus allowing us to solve problems in very different...