Description: This book is a true sense of in-depth learning introductory book, in-depth and shallow analysis of the principles of in-depth learning and related technologies. Using Python 3, the book tries not to rely on external libraries or tools, starting from basic mathematical knowledge, to lead readers to create a classic in-depth learning network from scratch, so that readers gradually understand in-depth learning in the process. The book not only introduces the basic knowledge of depth learning and neural network, such as the concept, characteristics, and so on, but also the error back propagation method, convolution neural network and so on. In addition, it also introduces the practical skills related to depth learning, the application of automatic driving, image generation, reinforcement learning and so on, and why to deepen the layer. In order to improve the accuracy of identification and other "why" problems.