April 2019
Intermediate to advanced
212 pages
5h 34m
English
OpenAI Gym tasks are available as environments that provide us with a state space and an action space, along with the rewards and outcome responses that we need to train a learning agent.
There are many environments to work with; some that are simpler to use and understand than others, and new environments are being added continually. We discussed many of these different environments in detail in the last chapter:

The preceding diagram is a still from the CartPole environment, a task that we'll be solving in Chapter 6, Digging Deeper into Deep Q-Networks with Keras and TensorFlow, using a deep Q-network.
The CartPole task ...
Read now
Unlock full access