Probabilistic Machine Learning: An Introduction
by Kevin Patrick Murphy.
MIT Press, 2021.
If you use this book, please be sure to cite
- Short table of
- Long table of
- Draft pdf file, CC-BY-NC-ND license. (Please cite the official reference below.)
- Python code
- Issue tracker. Use this to report problems with the book and/or code. Be sure to specify which
release (date stamp) of the book you are using. Please specify pdf and print page number (which sometimes differ). For reporting small typos, please collect
a batch of errors into a doc, and create a single issue (or add to an existing open issue list).
author = "Kevin P. Murphy",
title = "Probabilistic Machine Learning: An introduction",
publisher = "MIT Press",
year = 2021,
url = "probml.ai"
Table of contents
since first release on 2021-01-03.
(Visits to pml-book
"Kevin Murphy’s book on machine learning is a superbly written,
comprehensive treatment of the field, built on a foundation of probability theory.
It is rigorous yet readily accessible, and
is a must-have for anyone interested in gaining a deep understanding of machine learning."
-- Chris Bishop,
"This book is a clear, concise, and rigorous introduction to the foundations of machine learning.
It beautifully bridges between the "traditional" topics and the more "modern" deep learning methods,
creating a unifying framework that contextualizes both of them. It's the book I recommend for people who are new to the
field and want to obtain a comprehensive view of the core principles and methods."
-- Daphne Koller, Insitro/ Stanford.
"This is a remarkable book covering the conceptual,
theoretical and computational foundations of probabilistic machine learning,
starting with the basics and moving seamlessly to the leading edge of this field.
The pedagogical structure of the book is extremely useful for teaching. One of my favorite parts is
that most of the figures of the book have a link to the associated
(python, JAX, tensorflow) code that is used to generate them,
often with comparisons between the different computational ways of solving the problems."
-- Michael Brenner, Harvard/ Google.
"This book could be titled 'What every ML PhD student should know'.
If you master the material in this book, you will have an outstanding foundation for successful research in machine learning.”
-- Tom Dietterich, U. Oregon
- "There are many books on machine learning out there, but none gives
such a well-rounded, up-to-date, and comprehensive view of the
field as this one. We use this book as reference reading for our
students taking the advanced machine learning course at Oxford to
introduce them to fundamental as well as current topics in the
field. I'm amazed at the amount of work that went into this
book---which will surely be used by many to train the next
generation of machine learning experts."
-- Yarin Gal, U. Oxford
"This is a terrific resource for machine learning students and researchers.
If you want to understand the foundations of modern machine learning then this is the book to read.
The text is particularly strong at marrying classical ideas from statistics and probability with more modern concepts such as deep learning."
-- Padhraic Smyth, UC Irvine
- "My favorite machine learning book just received a face-lift!
'Probabilistic Machine Learning: An Introduction' is the most
comprehensive and accessible book on modern machine learning by a
It now also covers the latest developments in deep learning and
causal discovery. With this upgrade it will remain the reference
book for our field that every respected researcher needs to have
on their desk." -- Max Welling,
"Prof Murphy's 2012 book was a triumph, covering both basic material
and also the state-of-the-art. The new 'Probabilistic Machine
Learning: An Introduction' is similarly excellent, and includes new
material, especially on deep learning and recent developments. It
will become an essential reference for students and researchers in
probabilistic machine learning."
-- Chris Williams, U. Edinburgh
I would like to thank the following people for helping with this book.
- People who helped write some sections:
Si Yi Meng,
- Proof readers: Peter Cerno, John Fearns.
- People who have provided feedback on parts of the book:
- People who have helped with the code:
Gerardo Durán Martín,
- People who have helped with the figures:
Sandeep Choudhary, and others who are credited in the figure captions.