search
search

The Rigakubu News

Disclaimer: machine translated by DeepL which may contain errors.

Introduction to Physics and Deep Learning from the Basics"

Kenji Fukushima, Professor, Department of Physics / Norisho Katsura, Associate Professor, Department of Physics

Kenji Fukushima and Hosho Katsura (Author)
"Introduction to Physics and Deep Learning from the Basics"

Kagaku Joho Shuppan K.K. (2022)
ISBN 978-4-910558-07-3

In the same way that generative AI has made our lives more convenient in recent years, there are increasing opportunities for physics research to make use of such methods, especially deep learning.

In this book, the authors, who specialize in physics, explain deep learning from a physicist's point of view in an unusual textbook. It is not a novel book, however, but rather an introductory book for readers with no prior knowledge of physics or deep learning, allowing them to learn about both at the same time. The first half of the book begins with an overview of what "learning" is, followed by an introduction to the fundamentals of quantum mechanics and statistical mechanics as taught in the Department of Physics in the preparatory section, and then introduces various methods of constructing neural networks in the introductory section. The second half of the book, consisting of a practical section and an application section, details the non-linear regression of deep learning and its application to specific problems in modern physics, with demonstrations.

Readers interested in physics will find Ising models and entropy, familiar from undergraduate-level textbooks, in action in unusual ways. Readers with no knowledge of physics will also get a glimpse into the world of quantum mechanics and statistical mechanics through deep learning. Before beginning a course in quantum mechanics or statistical mechanics in the Faculty of Science, a light reading of this book is sure to deepen one's understanding of the subject. Another feature of this book is that it contains the rudiments of graph theory, which is rarely covered in introductory textbooks.


 

Faculty of Science News, September 2023

Science Bookshelf>