This book provides a systematic introduction to advanced techniques in deep learning and practical applications based on the MindSpore AI computing framework. The book is divided into 10 chapters, covering topics such as data processing, network construction, training and inference performance optimization, model security and privacy, model reliability, explainable AI, fairness issues in AI, data-driven AI models, solving scientific computing equations with AI, and accelerating scientific methods with AI. To assist readers in learning, the book also includes example code for advanced deep learning techniques implemented based on MindSpore.Combining MindSpore** open-source technology with the theoretical foundation of deep learning, this book expands MindSpore’s applicability. It can serve as a textbook for undergraduate and graduate students in AI, intelligent science and technology, computer science and technology, electronic information engineering, automation, and other majors. It is also suitable as a reference guide for software development engineers and researchers engaged in deep learning-related work.