Write a 1-2 page description of how you could use Q-learning (the reinforcement learning algorithm) to create a tic-tac-toe agent. Your description may include pseudo-code where appropriate (or describe things in English). Your description must include the following:

• A description of what your agent will do each time it is asked to make a move (how it uses the results of Q-learning to decide what to do).

• A general description of the software you would need in order to train the agent, and a plan for using this software.

• A detailed description of how/when Q values are changed. You should include specifics such as the range of Q values, initial Q values, etc. Pseudo code would be great!

• An estimate as to how well you think this would work (and why).

• Anything else you find interesting...