Technical editor: Zong en from sifou Office
SegmentFault has he reported the official account number: SegmentFault
In recent years, artificial intelligence technology, represented by machine learning and knowledge mapping, has gradually become popular. From license plate recognition, face recognition, speech recognition, intelligent question answering, recommendation system to automatic driving, people may use artificial intelligence technology consciously or unconsciously in daily life.
These technologies are inseparable from the long-term efforts of researchers in the field of artificial intelligence. Especially in recent years, due to the increase of data, the enhancement of computing power, the maturity of learning algorithms and the richness of application scenarios, more and more people begin to pay attention to this “new” research field: deep learning.
Deep learning takes neural network as the main model, which is used to solve the representation learning problem in machine learning. However, due to its powerful ability, deep learning is increasingly used to solve some general artificial intelligence problems, such as reasoning, decision-making and so on. At present, deep learning technology has achieved extensive success in the academic and industrial circles, and has been highly valued, and set off a new round of artificial intelligence boom.
“Nerve network and deep learning” is a free online book. It has a moderate demand for mathematical knowledge of readers, taking into account both theory and practice. The best solutions to many problems in the field of image recognition, language recognition and natural language processing are given in this paper.
This book will teach you: neural network, a wonderful physics inspired programming paradigm, can let the computer learn from observation data; deep learning, a powerful collection of many technologies in neural network learning.
The original address of this book is:
Michael Nielsen is a physicist, scientific writer and computer programming researcher.
Chapter 2, Chapter 3, Chapter 5, Chapter 6 and the appendix.
Freeman Zhang: translating the rest of the chapters, supplementing the missing contents, proofreading, sorting and typesetting.
Scanning the end of the two-dimensional code into WeChat official account SegmentFault backstage reply: “neural network”
You can get the latest version of the Chinese version of “neurology and deep learning” PDF learning document for free.
Part of the summary: