Math Behind Reinforcement Learning, the Easy Way