Название: Reinforcement Learning From Scratch: Understanding Current Approaches - with Examples in Java and Greenfoot
Автор: Uwe Lorenz
Издательство: Springer
Год: 2022
Страниц: 195
Язык: английский
Формат: pdf (true), epub
Размер: 46.5 MB

In ancient games such as chess or go, the most brilliant players can improve by studying the strategies produced by a machine. Robotic systems practice their own movements. In arcade games, agents capable of learning reach superhuman levels within a few hours. How do these spectacular Reinforcement Learning (RL) algorithms work? With easy-to-understand explanations and clear examples in Java and Greenfoot, you can acquire the principles of Reinforcement Learning and apply them in your own intelligent agents. Greenfoot and the Hamster model are simple but also powerful didactic tools that were developed to convey basic programming concepts.

The result is an accessible introduction into Machine Learning (ML) that concentrates on Reinforcement Learning. Taking the reader through the steps of developing intelligent agents, from the very basics to advanced aspects, touching on a variety of Machine Learning algorithms along the way, one is allowed to play along, experiment, and add their own ideas and experiments.

Reinforcement Learning is an area of Machine Learning and belongs to the broad field of Artificial Intelligence (AI). Many believe that reinforcement learning has something to do with artificial neural networks. Although the book also deals with reinforcement learning algorithms that use neural networks, this does not play a central role. Deep reinforcement learning is actually a subset of reinforcement learning, where some functions are implemented with deep neural networks. It may surprise that explanations of the “Asynchronous Advantage Actor-Critic (A3C)” or “Proximal Policy Optimization (PPO)” are possible without the use of neural networks.

Regarding the quotation above, perhaps the content in this book is not quite reduced to a beginner’s level, but it should be very suitable, especially for newcomers to Machine Learning. It is a book for those with some basic knowledge of programming and high school level math. It is useful, for example, for continuing education for teachers and instructors who want to gain insight into programming adaptive agents. The book should also be appropriate for technicians, computer scientists, or programmers who want to better understand RL algorithms by studying and implementing learning these algorithms from scratch themselves (especially if they have been socialized with Java) or for students who want to study Machine Learning and intelligent agents.

Contents:
1. Reinforcement Learning as a Subfield of Machine Learning
2. Basic Concepts of Reinforcement Learning
3. Optimal Decision-Making in a Known Environment
4. Decision-Making and Learning in an Unknown Environment
5. Artificial Neural Networks as Estimators for State Values and the Action Selection
6. Guiding Ideas in Artificial Intelligence over Time