IBM Research Africa ‚ÄčReinforcement Learning Fundamentals