Deep Learning Master's Notes Volume 1: Basic Algorithms

Through solid, detailed content and clear structure, this book analyzes and introduces deep learning algorithms from the aspects of algorithm theory, algorithm source code, and experimental results. This book consists of three parts. The first part mainly introduces some convolutional neural networks of deep learning in the direction of computer vision, starting from three directions: basic backbone network, lightweight CNN, and model architecture search, and introduces milestone algorithms in the direction of computer vision. The second part mainly introduces the important breakthroughs of deep learning in the direction of natural language processing, including basic sequence models and model pre-training. The third part mainly introduces the progress of deep learning in model optimization, including model optimization methods. By reading this book, readers can have an in-depth understanding of the mainstream basic deep learning algorithms, build their own knowledge system, understand the essence of the algorithm, and learn model optimization methods. Both teachers and students engaged in deep learning research, as well as staff engaged in algorithm implementation, can benefit from this book.