https://trunghng.github.io/posts/reinforcement-learning/func-approx/