Hands-On Q-Learning with Python

Creating a baseline agent

In this section, we'll implement an agent that takes random actions and does not keep track of its actions or learn from them. We'll get started on building an actual Q-learning algorithm in Chapter 4, Teaching a Smartcab to Drive Using Q-Learning. For now, all your agent will be able to do is to take random actions.

As part of our analysis, we'll be comparing the success of this randomly-acting agent to the results of an optimized Q-learning agent. The randomly-acting agent is called our baseline agent, and we will use it as a control to which we'll compare the performance of future machine learning models. We'll discuss the significance of baseline models at the end of the chapter.