The StatQuest Illustrated Guide to Neural Networks and AI: With hands-on examples in PyTorch!!! (available from Amazon through the StatQuest store) strikes an excellent balance between accessibility and technical depth. Josh Starmer, PhD, builds on his previous work while making neural networks approachable for both students and practitioners. This book has a similar feel and vibe to the previous book, The StatQuest guide to Machine Learning.
What makes this book particularly valuable is its comprehensive coverage of architectures relevant to real-world Earth Science applications. The sections on Long Short-Term Memory networks (LSTMs) are essential for anyone working with time series data — like forecasting streamflows in hydrology. Similarly, the coverage of Convolutional Neural Networks (CNNs) provides fundamental knowledge for working with image data, from satellite imagery classification to computer vision applications.
Each chapter comes with practical, modern PyTorch exercises that bridge theory and implementation. The book progresses logically from fundamental concepts through to advanced topics like transformers, making it accessible for those transitioning from traditional statistical methods to deep learning approaches. Starmer's explanations of concepts like backpropagation, cross-entropy, and attention mechanisms are clear and practical, enhanced by his fun illustration style that helps build intuition.
For readers seeking to deepen their mathematical understanding, two excellent free online resources complement Starmer's work. Mathematics for Machine Learning by Marc Peter Deisenroth, A. Aldo Faisal, and Cheng Soon Ong provides rigorous coverage of the underlying mathematical concepts, while An Introduction to Statistical Learning with Python by Gareth James, Daniela Witten, Trevor Hastie, Robert Tibshirani, and Jonathan Taylor offers a comprehensive foundation in statistical and machine learning methods. These resources are particularly valuable for students and professionals who want to further understand the mathematical principles behind neural network architectures and optimization techniques.
Thomas Martin is an AI/ML Software Engineer at the NSF Unidata Program Center. Have questions? Contact support-ml@unidata.ucar.edu or book an office hours meeting with Thomas on his Calendar.
