https://medium.com/@monocosmo77/how-bellman-operators-work-part1-advanced-reinforcement-learning-6b595dd6cf47
