This book provides a comprehensive exploration of both classical and modern models in deep learning, with a primary focus on the theory and algorithms that underpin the field. It delves into the fundamental concepts of neural networks, helping readers understand the key principles that guide the design and application of neural architectures across various domains. Central to the book are critical questions such as: Why do neural networks work? When do they outperform traditional machine-learning models? What makes depth in networks useful, and what are the challenges in training them? It also examines the pitfalls and complexities involved in deep learning, providing insights into the nuances of building and optimizing neural networks. The book also emphasizes the practical application of deep learning techniques. It offers in-depth discussions on how neural architectures are tailored for different types of data, including text, images, and graphs. By exploring a wide range of applications, the book helps practitioners gain a deeper understanding of how to design neural networks for diverse problems. The content is organized into three categories, ensuring a well-rounded and systematic approach to learning deep learning techniques.