Reinforcement learning is not a trivial topic and even from a more practical perspective, mastering the subject requires some background in computer programming, math and probabilities. Although there's a increasing number of libraries which offers environments and algorithms out-of-the-box, a ground base on reinforcement learning theory is essential to choose the appropriate algorithms for each kind of problem and to tune their hyperparameters when it's necessary.
In this post, we bring 5 courses developed by top of mind universities and companies, and you can watch them online for free. Let's check them:
1. Reinforcement Learning by Georgia Tech (Udacity)
This course was developed by Georgia Tech and it's available for free on the Udacity online platform. All you have to do is create your account and sign in for the classes. The course covers the initial concepts of reinforcement learning, including temporal difference learning (TD learning), Markov decision process (MDP) and topics on game theory. You access the course in this link:
Reinforcement Learning by Georgia Tech
2. Reinforcement Learning Explained (edX)
This course is part of the Microsoft Professional Program in Artificial Intelligence, but you can take it alone. The course covers basic and advanced topics like Policy Gradient and Actor Critic. The course is free, but you can pay for a verified certificate. Below is the link for the course:
Reinforcement Learning Explained
3. Practical Reinforcement Learning (Coursera)
This course is part of the Advanced Machine Learning Specialization, offered by the Russia's Higher School of Economics. It's a 4 week course which covers topics like value/policy iteration, q-learning, policy gradient and advances to deep reinforcement learning with Deep Q-Learning. The course is paid if you want a certificate and access the assignments, but you can audit the course for free.
Practical Reinforcement Learning
4. Introduction to Reinforcement Learning
This course was lectured by prof. David Silver at the UCL (London's Global University). The course covers topics like Markov Decision Process, Dynamic Programming, Value Function Approximation, Policy Gradient, Exploit-Exploration Dilemma and others. You can access the lectures at the DeepMind's Youtube Channel as well as the lectures slides at the course website.
Introduction to Reinforcement Learning, by prof. David Silver
5. Deep Reinforcement Learning (CS 294-112)
The last course of our list was offered by the UC Berkeley. It covers from the basics of Reinforcement Learning to advanced topics. You can find links to slides, discussion groups at the course front-page and the recorded lectures at the CS 294-112 Youtube Channel.
Deep Reinforcement Learning (CS 294-112)
The courses listed above covers a wide range of topics on Reinforcement Learning and gives you all the theory necessary to start developing your own intelligent agents, either they are intended to play Atari games, stock trading or build robots. If you have any questions about the courses or reinforcement learning, don't hesitate to write at rodolfo@reinforcementlearning4.fun.
Originally published at: Top 5 Free Courses in Reinforcement Learning
Top comments (0)