Reinforcement learning is a ML training method based on rewarding desired behaviours and punishing undesired ones. A reinforcement learning agent can perceive and interpret its environment, take actions and learn through trial and error. Reinforcement learning is largely used in autonomous driving, automated cooling for data centres, recommendation engines, personalised chatbots, stock trading etc.
Here, we look at the top resources to learn reinforcement learning in 2022:
RL course by David Silver
DeepMind research lead David Silver’s course on reinforcement learning taught at University College London is laid out in ten YouTube videos. The videos cover Introduction to Reinforcement Learning, Markov Decision Processes, Planning by Dynamic Programming, Model-Free Prediction, Model-Free Control, Value Function Approximation, Policy Gradient Methods, Integrating Learning and Planning, Exploration and Exploitation, Case Study: RL in Classic Games. To access slides, assignments, exams, check out the link.
Introduction to Reinforcement Learning with Function Approximation
Rich S. Sutton, a research scientist at DeepMind and computing science professor at the University of Alberta, explains the underlying formal problem like the Markov decision processes, core solution methods, dynamic programming, Monte Carlo methods, and temporal-difference learning in this in-depth tutorial.
A History of Reinforcement Learning
Prof AG Barto, professor emeritus of computer science at University of Massachusetts Amherst, offers a detailed lecture. The chapters include “Hedonistic neuron” hypothesis, Supervised learning, Reinforcement learning, A unique property of RL, Edward L Thorndike, Law of Effect, RL= Search+Memory, Our first surprise, Though there were exceptions, An early paper with Rich Sutton, Associative Memory Networks, Associative Search Network and many more.
The course entails a series of lectures by Prof Balaraman Ravindran, Computer Science and Engineering and Robert Bosch Centre for Data Science and AI, IIT-Madras on Reinforcement Learning. The course introduces the basic mathematical foundations of reinforcement learning and highlights some of the recent directions of his research. The 12 weeks lecture contains preparatory material, introduction to RL and immediate RL, Bandit Algorithms, Policy Gradient Methods and introduction to Full RL, MDP Formulation, Bellman Equations & Optimality Proofs, Dynamic Programming & Monte Carlo Methods, Monte Carlo & Temporal Difference Methods, Eligibility Traces, Function Approximation, DQN, Fitted Q & Policy Gradient Approaches, Hierarchical Reinforcement Learning, Hierarchical RL: MAXQ and POMDPs.
Artificial Intelligence: Reinforcement Learning in Python
Artificial Intelligence: Reinforcement Learning in Python is a complete guide to reinforcement learning with stock trading and online advertising applications. The 14.5 hours-course is available as on-demand video in Udemy. The guide will teach you to apply gradient-based supervised machine learning methods to reinforcement learning, understand reinforcement learning on a technical level, understand the relationship between reinforcement learning and psychology, and implement 17 different reinforcement learning algorithms.
Reinforcement Learning in Unity
The students can learn to set up reinforcement learning in Unity3D and unlock the power of combining game engines with artificial intelligence by using it to train a tile to balance a little ball. Details can be found in https://github.com/Unity-Technologies
Introduction to reinforcement learning
Practical Reinforcement Learning
Practical Reinforcement Learning by Coursera covers the foundations of RL methods: value/policy iteration, q-learning, policy gradient, etc; using deep neural networks for RL tasks; state of the art RL algorithm; and teaching neural networks to play games.
Deep Reinforcement Learning
The course on GitHub has a series of articles and videos to help you master the skills and architectures to become a deep reinforcement learning expert.The course will help build a strong professional portfolio by implementing agents with Tensorflow and PyTorch that learn to play Space invaders, Minecraft, Starcraft, Sonic the Hedgehog and more.