Reinforcement Learning introduction via Multi arm bandit

If it meets the guidelines, this might make a good 'Show HN'. Show HN guidelines: https://news.ycombinator.com/showhn.html