January 2020
Intermediate to advanced
346 pages
9h 8m
English
In this chapter, we will demonstrate how genetic algorithms can be applied to reinforcement learning—a fast-developing branch of machine learning that is capable of tackling complex tasks. We will do this by solving two benchmark environments from the OpenAI Gym toolkit. We will start by providing an overview of reinforcement learning, followed by a brief introduction to OpenAI Gym, a toolkit that can be used to compare and develop reinforcement learning algorithms, as well as a description of its Python-based interface. Then, we will undertake two Gym environments, MountainCar and CartPole, and develop genetic algorithm-based programs to solve the challenges they present.
In this chapter, we ...
Read now
Unlock full access