Q-Learning, what’s that?
Well to make it short it’s a very simple method of reinforcement learning, which again is a whole category of machine learning methods based on the idea of rewarding desired behaviour of an agent and/or punishing undesired one to make it learn.
Still too abstract? I made a series of three short videos for university, explaining reinforcement learning and later Q-Learning in more detail. Alternatively you could just search it on the web, there are probably better explanations.
In this project I implemented Q-Learning and applied it to two applications.
The Cart Pole Balancing Problem
This application is part of gym, a standard API for reinforcement learning in python.
The premise is simple: There is a cart with a pole on top and we need to keep it balanced by moving the cart left and right.
For details of the implementation and results you can check out the following links:
The Open Race Car Simulator (TORCS)
According to their website, TORCS is “ a highly portable multi platform car racing simulation” that “is used as ordinary car racing game, as AI racing game and as research platform”.
While all that is probably true, I have to say that it can be a bit tricky to set up and get everything running, especially if you want to use python as client language.
Again, I’ll spare you (and me) the details and leave you two links for details: