New Edition of 'Hands-On Machine Learning with Scikit-Learn, PyTorch' Released
October 14, 2026 · 2 min read
A new edition of 'Hands-On Machine Learning with Scikit-Learn, PyTorch' has been published by Hanbit Media on September 28, 2026. This updated version, authored by Aurélien Géron and translated by Park Hae-sun, replaces the previous edition's focus on TensorFlow and Keras with a complete rewrite centered on PyTorch.
The book expands its coverage to include three new chapters on transformers, enhancing its scope to natural language processing, vision, and multimodal applications. It also features a practical guide to fine-tuning pre-trained models and large language models (LLMs).
Yongpung Bookstore's technical section MD, Lee Won-jun, remarked that this book provides a comprehensive map of machine learning, from classic methods in Scikit-Learn to diffusion models and reinforcement learning. He noted the consistency in terminology across both this book and a previous work on deep learning by the Keras creator, making it easier for readers to transition between the two.
The book consists of 1,112 pages, which may pose a challenge for some readers. Book team leader Kwon Ki-won suggested a reading order, recommending beginners start with the first part on end-to-end projects using Scikit-Learn. For those familiar with the third edition, he advised skimming the first part and moving directly to the foundational PyTorch concepts and the new chapters on transformers.
Developers already using LLMs can also begin with the chapter on fine-tuning pre-trained models, allowing them to find relevant concepts as needed. This edition serves both as a textbook for comprehensive study and a reference for specific topics as challenges arise.
The publisher emphasizes the importance of understanding AI's creation, not just its application, positioning this edition as a broad starting point for developers looking to solidify their knowledge of underlying learning principles in a rapidly evolving field.