Current location - Training Enrollment Network - Books and materials - What is the basis of deep learning?
What is the basis of deep learning?
If you want to take the direction of deep learning in the future, in fact, the courses of undergraduate mathematics department are basically enough. If it is not enough, read some materials and books. You don't need to be a graduate student in mathematics.

I recommend the book Deep Learning. The authors are Ian Goodfellow, Yue Shu Bengio and Aaron courville. The Chinese version of this book was published on July 22nd, 20th17th. This book was written by many translators. The book Deep Learning introduces the basic knowledge of mathematics, the experience of machine learning and the theory and development of deep learning at this stage. Whether it is an artificial intelligence technology enthusiast or a related practitioner, using this book is very beneficial. In addition, if readers want to be familiar with some mathematical knowledge, this book also makes some introductions, including matrix, derivative and other basic contents. Readers can read from beginning to end.

A major feature of the book "Deep Learning" is that it introduces the essence of deep learning algorithm and gives the logic behind the algorithm. Without specific code implementation, people who don't write code can understand it. In order to make it easier for readers to read, the author specially draws the content organization chart of this book, and points out the relationship between the contents of the 20 chapters of this book. Readers can choose to read at will according to their own background or needs.

The link of the official Chinese version of GitHub is as follows: /exa city/deep learning book- Chinese.

Deep learning

JD.com

15 1.20

buy

In addition, there is a book, Deep Learning by Hands. This book aims to provide readers with an interactive learning experience about deep learning. This book not only expounds the algorithmic principles of deep learning, but also demonstrates their implementation and operation. Different from traditional books, each section of this book is a downloadable and operable Jupyter notebook, which integrates words, formulas, images, codes and operation results. In addition, readers can also visit and participate in the discussion of the contents of the book.

The book is divided into three parts: the first part introduces the background of deep learning and provides preparatory knowledge, including the concept and technology of deep learning foundation; The second part describes the important components of deep learning calculation, and also explains the convolutional neural network and cyclic neural network that have made deep learning successful in many fields in recent years. The third part evaluates the optimization algorithm, investigates the important factors that affect the computing performance of deep learning, and lists the important applications of deep learning in computer vision and natural language processing respectively.

This book also covers the methods and practices of deep learning, mainly for college students, technicians and researchers. Reading this book requires readers to understand the basics of Python programming or the basics of linear algebra, differentiation and probability described in the appendix.

Practical learning and deep learning

JD.com

76.50

buy

If you feel that your knowledge of mathematics is not enough, you can read the book Mathematics for Deep Learning. This book introduces the mathematical knowledge related to deep learning in a simple way with rich illustrations and concrete examples. Chapter 1 introduces the general situation of neural network; Chapter 2 introduces the basic mathematical knowledge needed to understand neural network; The third chapter introduces the optimization of neural network; In chapter 4, neural network and error back propagation method are introduced. Chapter 5 introduces deep learning and convolutional neural networks. Excel is used for theoretical verification in the book to help readers intuitively experience the principle of deep learning.