Ali Ghodsi, professor in the Department of Statistics and Actuarial Science, has co-authored a new textbook, Elements of Deep Learning.
Co-authored with Benyamin Ghojogh, the book offers a comprehensive introduction to deep learning and neural networks, combining mathematical rigor with hands-on practice. Covering both foundational concepts and emerging topics, the text explores areas including convolutional neural networks, Transformers, large language models, diffusion models, graph neural networks, reinforcement learning, and more.
Designed for advanced undergraduate and graduate students, instructors, researchers, and professionals, Elements of Deep Learning emphasizes practical implementation through PyTorch-based code examples, design insights, and real-world applications across vision, language, signal processing, healthcare, and related fields.
As Ghodsi explains, the book “balances mathematical rigour with hands-on practice, offering clear, step-by-step explanations grounded in theory and real-world application,” helping readers translate deep learning concepts into working systems.
The book is organized into five sections; fundamentals, sequence models, generative models, emerging topics, and practice, providing readers with a unified roadmap for understanding modern deep learning.
You can learn more about the book on the publisher’s website and learn more about deep learning through Professor Ghodsi's teaching videos.
