Interpretable Machine Learning with Python: Build explainable, fair, and robust high-performance models with hands-on, real-world examples Paperback – 31 Oct. 2023

Today, interpretability is more important than ever. Consequently this book comes at just the right time.Interpretable Machine Learning with Python is a comprehensive (approximately 600 page) book, and with a foreword by renowned author Denis Rothman, you know you are in for something special.The book does not disappoint. Starting with why interpretability is important, key concepts, and the challenges of interpretation, the book is written in an easy to understand style with lots of code and examples. The middle chapters go through some common techniques such as LIME and Anchors, with the book ending on some advanced topics such as monotonic constraints and adversarial robustness. However, this reader found the chapter on interpreting NLP transformers to be invaluable. Each of the topics (visualising attention, interpretating token attributions, and the Learning Interpretability Tool) is covered in detail, with Python examples and detailed explanations.In summary, this book is invaluable for data professionals, data scientists, and other professionals who work in AI. Highly recommended.