The best machine learning books, according to hackers
10 machine learning books recommended in Hacker News comments over the past year, ranked by how many separate people recommend each one.
- Most recommendedBuild a Large Language Model (From Scratch) by Sebastian Raschka16 people recommend it
- Free and lovedThe Little Book of Deep Learning by François Fleuret3 people recommend it
All machine learning books
- Build a Large Language Model (From Scratch) by Sebastian RaschkaPeople recommend it for clearly explaining how LLMs work by having readers build a GPT-style model step by step in PyTorch. Critics raised no complaints.16recommend
- Build an LLM from Scratch by Sebastian RaschkaReaders say working through it built a real understanding of transformers and backpropagation, and the author made it free to read online.8recommend
- Python Data Science Handbook by Jake VanderPlasA well-written intro to data science and machine learning in Python that covers a lot of ground; the caveat raised is that it is more an introduction than depth.8recommend
- The Little Learner by Daniel P. Friedman and Anurag MendhekarQuick, easy, project-based introduction to deep learning that builds from primitives up; critics say it is not a good first book for complete beginners.6recommend
- Probabilistic Machine Learning by Kevin MurphyPeople recommend it as a thorough prerequisite for understanding the ideas behind modern deep learning, with well-regarded volumes covering deep learning and reinforcement learning.5recommend
- The Welch Labs Illustrated Guide to AI by Welch LabsPeople recommend it as a visual introduction to AI from the team behind the Welch Labs videos, and call it a beautiful book that is pleasant to read.5recommend
- Pattern Recognition and Machine Learning by Christopher BishopPeople recommend it as a useful and informative book on pattern recognition and machine learning, good enough to keep on a bookshelf.3recommend
- The Little Book of Deep Learning by François FleuretRecommended as a free high-level introduction to deep learning concepts, a useful start for software developers. Main caveat: it stays high-level and somewhat handwavy.3recommend
- Deep Learning: Foundations and Concepts by Chris BishopRecommended for building intuition and tools for machine learning fundamentals, and a free online version is available on the publisher's site.3recommend
- Advances in Financial Machine Learning by Marcos Lopez de PradoPeople recommend it as a deep dive into machine learning techniques for financial modeling and data-driven trading. Few critics weighed in, and none raised a recurring complaint.2recommend