Exploring the Learning Algorithm Landscape - DDPG (Actor-Critic), PPO (Policy-Gradient), Rainbow (Value-Based)

In the previous chapter, we looked at several promising learning environments that you can use to train agents to solve a variety of different tasks. In Chapter 7, Creating Custom OpenAI Gym Environments – CARLA Driving Simulator, we also saw how you can create your own environments to solve the task or problem that you may be interested in developing a solution for, using intelligent and autonomous software agents. That provides you with directions on where you can head after finishing in order to explore and play around with all the environments, tasks, and problems we discussed in this book. Along the same lines, in this chapter, we will discuss several promising learning algorithms that serve as future references for your intelligent agent development endeavors...