Reinforcement Learning - Developing Intelligent Agents

Solving Cart and Pole with Deep Q-Network - Reinforcement Learning Code Project

Welcome back to this course on reinforcement learning! In this episode, we'll discuss how we can tune our current code project in order for our deep Q-network to solve the Cart and Pole environment.

We last left off having completed our program by training our deep Q-network on the Cart and Pole environment. There, we saw that the network's average reward consistently grew over the 1000 episode training duration, with its final score reaching 100-episode average reward of 90.