Chapter 4: Reinforcement Learning in the Real World – Building Cryptocurrency Trading Agents
Deep reinforcement learning (deep RL) agents have a lot of potential when it comes to solving challenging problems in the real world and a lot of opportunities exist. However, only a few successful stories of using deep RL agents in the real world beyond games exist due to the various challenges associated with real-world deployments of RL agents. This chapter contains recipes that will help you successfully develop RL agents for an interesting and rewarding real-world problem: cryptocurrency trading. The recipes in this chapter contain information on how to implement custom OpenAI Gym-compatible learning environments for cryptocurrency trading with both discrete and continuous-value action spaces. In addition, you will learn how to build and train RL agents for trading cryptocurrency. Trading learning environments will also be provided.
Specifically, the following recipes will be...