Hands-On Deep Learning with R

Reinforcement learning example

In this example, we will create a sample environment for reinforcement learning. The concept of reinforcement learning will be explored in more detail in a later chapter. For this example, we will generate a series of states and actions, along with the reward for taking those actions, that is, whether taking the action led to the desired result or a negative consequence. Afterward, we will define how our agent should respond or learn from actions. Once all of this has been defined, we will run the program and the agent will navigate through the environment to learn to solve the task. We will define and run a minimal reinforcement learning example by running the following code:

library(ReinforcementLearning) ...