Skip to content

📚 Recommended Books in Machine Learning / Deep Learning

This is a curated list of foundational and influential books in the field of machine learning and deep learning, along with personal notes and commentary.

Core Machine Learning Books

Pattern Recognition and Machine Learning

Author: Christopher Bishop
- Widely referred to as PRML, this is one of the most comprehensive treatments of machine learning techniques such as neural networks, graphical models, and boosting. - Strongly Bayesian in approach. Treats bias-variance as a “frequentist illusion.” - Considered tough but highly rewarding. Especially insightful if you do the exercises. - Not focused on deep learning; for that, see the Goodfellow/Bengio/Courville book.

Pattern Classification (3rd Edition)

Authors: R. Duda, P.E. Hart, D.G. Stork
- Known informally as Duda and Hart. - Terse and technical; originally the “bible” of pattern classification. - Slightly outdated but still valuable, especially in its treatment of linear classifiers. - Includes updated computer exercises in the 3rd edition.

Machine Learning

Author: Tom Mitchell
- More concise and beginner-friendly than PRML or Duda & Hart. - Covers concept learning, decision trees, and neural networks. - Highly readable and still insightful, though less aligned with modern deep learning practices.

The Master Algorithm

Author: Pedro Domingos
- A popular science-style book offering a sweeping overview of ML philosophies. - Great for understanding different schools of thought in ML.

📖 Books I've Heard Are Excellent

  • The Elements of Statistical Learning – Hastie, Tibshirani, Friedman
  • Bayesian Reasoning and Machine Learning – David Barber
  • Machine Learning: A Probabilistic Perspective – Kevin Murphy
  • Information Theory, Inference and Learning Algorithms – David MacKay
  • Deep Learning – Ian Goodfellow, Yoshua Bengio, Aaron Courville
    -- The only book here directly focused on deep learning. See my impression here.

Historical / Advanced References

  • Perceptrons: An Introduction to Computational Geometry – Marvin Minsky & Seymour Papert - Historically significant; challenged early neural net developments.

  • Parallel Models of Associative Memory – Geoffrey Hinton - Of historical interest and influence.