We've gone over multi-armed bandit problems in detail and discussed practical ways in which they can be applied to real-world problems such as advertising and product testing. We've introduced different approaches to solving the problem and suggested opportunities for further research into each of these approaches.
This is only an introduction to the multi-armed bandit problem space, which is well worth researching further and has many exciting applications to explore.
In the next chapter, we'll explore further the types of problems we can solve using our knowledge of Q-learning, including the additional environments offered by OpenAI Gym. We'll leave you with ideas for future projects to develop your skills as a RL practitioner and researcher and you'll be familiar with many of the domains in which you can practice your skills.
...