The third part in the Marty Learns to Swing series. This time out Marty pursues Q-Learning but finds that an Exhaustive approach suits him better.
Marty Learns to Swing – Part 2
In this, the second part of the series, I develop an OpenAI Gym to model Marty the Robot on a swing and get it to the point where we can start testing strategies for keeping Marty swinging once he gets started.