Introduction to Reinforcement Learning — Machines Learn Like Humans
Introduction to Reinforcement Learning — Machines Learn Like Humans To some, reinforcement learning is the most “real” artificial intelligence among other machine learning / deep learning methods. It studies the mathematical and algorithmic basis for how machines can learn to do anything, just by themselves without any instructions. The study gets even more exciting when we learn that some of the concepts we discovered in reinforcement learning, for example “temporal difference learning” (TD), has correspondence in human brain [1]. In this article, you’ll be introduced with the fundamental concepts of reinforcement learning both with intuition and mathematical basis. What is Reinforcement Learning? Although, you’ve been introduced with a brief definition of reinforcement learning, that definition does not cover the whole story. In essence, reinforcement learning studies an intelligent being making decisions [2], Under uncertainty With delayed rewards and penalties Trying to obtai...