Summary
Reinforcement learning is posed to play a significant role in automation. Smart city is a great field to leverage the power of RL. In this chapter, we discussed three sample applications: Traffic light control, ancillary service provision by electricity consuming appliances, and detecting cyberattacks in a power grid. The first problem allowed us to showcase a multi-agent setting, we used a price-like indirect control mechanism for the second one, and the last one was a good example of advanced input preprocessing in partially observed environments.
In the next and the last chapter, we will wrap up the book with a discussion on the challenges of real-life reinforcement learning and related future directions.
