book1.html

<h1>Probabilistic Machine Learning: An Introduction</h1>
by <a href="https://www.cs.ubc.ca/~murphyk/">Kevin Patrick Murphy</a>.
<br>
	MIT Press, February  2022.
	
<p>
<img src="pml1/TOC/cover1.jpg"
     alt="Book cover"
     style="height:200;">
<p>
	
	
<h2>Key links</h2>

<ul>

  <li> <a href="pml1/TOC/toc1.pdf">Long table of
      contents</a>

  <li> <a href="pml1/TOC/preface1.pdf">Preface</a>

  <li> <a href="https://github.com/probml/pml-book/releases/latest/download/book1.pdf">
    Draft pdf file</a>.  CC-BY-NC-ND license. (Please cite the official reference below.)

	  <li> <a href="https://www.amazon.com/Probabilistic-Machine-Learning-Introduction-Computation/dp/0262046822/ref=tmm_hrd_swatch_0?_encoding=UTF8&qid=&sr=">
		  Pre-order hardcopy from Amazon</a> (expected release date: February 2022).
	  
	<li> <a href="https://github.com/probml/pml-book/tree/main/pml1/figures/images">All the figures</a> (png files)
		
    <li> <a href="https://github.com/probml/pml-book/tree/main/pml1/figure_notebooks">Code to reproduce all the figures</a> 
      
	        <li> <a href="https://github.com/probml/pml-book/tree/main/pml1/README.md">Supplementary material</a> 
			
  <li> <a href="https://github.com/probml/pml-book/issues">Issue tracker</a>. 
	  
	<li><a href="https://groups.google.com/u/1/g/probml-book/">Mailing list for major announcements</a> (Low traffic)
	  
	  
	<li> <a href="#endorsements">Endorsements</a>
</ul>

If you use this book, please be sure to cite
<pre><code>
 @book{pml1Book,
 author = "Kevin P. Murphy",
 title = "Probabilistic Machine Learning: An introduction",
 publisher = "MIT Press",
 year = 2022,
 url = "probml.ai"
}
</code></pre>



  
<h2>Table of contents</h2>

    <img src="pml1/TOC/toc1-short.png"
     alt="TOC 2021-07-20"
     style="height:700;">


<h2><a id="code">Code accompanying the book</h2>

	Code to recreate all the figures can be found in a series of colabs, one per chapter,
stored  <a href="https://github.com/probml/pml-book/tree/main/pml1/figure_notebooks">
  here</a>.
When reading the pdf version of the book, you can  click on any link labeled <b>figures.probml.ai/x.y</b> and it will open up the  colab for chapter x;
	the cursor should  scroll down to the cell for figure y. 
	Once you get there, 
	click on the button labeled 'setup' and it will install any necessary code.
(The first time you do this it may take about 10 seconds, but subsequent setups for other cells in the same chapter should be faster,
	even if they open in a new tab.)
	After setup, click on the following cell and it will  run the code for you.
	(It should automagically install any missing packages as well, although you may need to run the cell twice to make this work.)
	
	<p>
	  The code for most figures is stored in individual files in the 
	  <a href="https://github.com/probml/pyprobml/tree/master/scripts">scripts</a> directory.
	  You can run these locally (on your laptop), but it's often faster to run in colab (especially for demos that use a GPU).
		To do this, just type `%run foo.py`. You can also edit the file in colab, and then rerun it. 
Note, however, that  changes to local files will not be saved beyond the current colab session.
(A better, but more complex, approach is to use VScode to ssh into the colab machine,
		see <a href="https://github.com/probml/probml-notebooks/blob/main/markdown/colab_gcp_tpu_tutorial.md">this page
		</a> for details.)
		
	<p>
There are also some inline links to code in the body of the book, labeled <b>code.probml.ai/foo</b>; these refer to 
demos that are not associated with any figure. Clicking on these links behaves in a similar way to the figure code (opening a tab for the appropriate
		colab cell).
In addition to the above, each chapter has
<a href="https://github.com/probml/pml-book/tree/main/pml1/supplements">supplementary code / material</a>.
		These will continue to be updated even after the book is published (contributions welcome!).

  
<p>
<h2><a id="endorsements"></a>Endorsements</h2>

<ul>

    <li> 
  <img src="https://img.shields.io/github/downloads/probml/pml-book/total"
       alt="download stats shield">


  <p>
	  <li>
		  "The deep learning revolution has transformed the field of machine learning over the last decade.
		  It was inspired by attempts to mimic the way the brain learns but it is grounded in basic principles of statistics, 
		  information theory,
decision theory and optimization. This book does an excellent job of explaining these principles and describes many of the "classical"
		  machine learning methods that make use of them. It also shows how the same principles can be applied in deep learning systems 
		  that contain many layers of features. 
		  This provides a coherent framework in which one can understand the relationships and tradeoffs between many different ML approaches,
		  both old and new." -- <a href="https://www.cs.toronto.edu/~hinton/">Geoff Hinton</a>. U. Toronto/ Google.
		  <p>
			  
    <li>
      "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."
	    -- <a href="https://www.microsoft.com/en-us/research/people/cmbishop/">Chris Bishop</a>,
	Microsoft Research.
	    
	    <p>
		    <li>
	    "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."
			    -- <a href="https://ai.stanford.edu/~koller/">Daphne Koller</a>, Insitro/ Stanford.


<p>
	<li>
		"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."
		-- <a href="https://www.seas.harvard.edu/brenner/Home.html">Michael Brenner</a>, Harvard/ Google.
		
	  <p>
		  <li>
	  "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.”
			  -- <a href="http://web.engr.oregonstate.edu/~tgd/">Tom Dietterich</a>, Oregon State U.
			  
			  <p>
			  <li>
"This book delivers a wonderful exposition of modern and traditional machine learning approaches through the language and lens of probabilistic reasoning. 
				  As such, it provides extremely valuable and much needed coherence, generalisation, and mathematical rigour, 
				  allowing the reader to gain a deep understanding of the interconnectivities between many different areas of data science, 
				  and so deliver more effective outcomes.
				  I highly recommend this second edition"
				  -- <a href="https://www.turing.ac.uk/people/researchers/mark-briers">Mark Briers</a>, Alan Turing Institute.

    <p>
    <li> "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."
      -- <a href="http://www.cs.ox.ac.uk/people/yarin.gal/website/">Yarin    Gal</a>, U. Oxford

	    
	  <p>
	    <li>
	      "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."
	      -- <a href="https://www.ics.uci.edu/~smyth/">Padhraic Smyth</a>, UC Irvine
		    
		          <p>
  <li> "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
    large margin.  
    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."  -- <a href="https://staff.fnwi.uva.nl/m.welling/">Max Welling</a>,
    U. Amsterdam 
	 
		    
      <p>
	<li>
	  "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."
	  -- <a href="https://homepages.inf.ed.ac.uk/ckiw/">Chris Williams</a>, U. Edinburgh

	  


    </ul>


  

  <p>
<h2><a id="ack"></a>Acknowledgements</h2>

I would like to thank the following people for helping with this book.

<ul>

<li> People who helped write some sections for vol 1:
	Sami Abu-El-Haija,
	Mathieu Blondel,
	Ines Chami,
  Krzysztof Choromanski,
	Roy Frostig,
  Justin Gilmer,
  Zico Kolter,
  Frederick Kunster,
  Lihong Li,
  Si Yi Meng,
  Aaron Mishkin,
  Byran Perozzi,
  Colin Raffel,
  Mark Schmidt,
  Sharan Vaswani,
  Andrew Wilson.
  
      <li> Proof readers: Peter Cerno, John Fearns.

      <li> People who have provided feedback on parts of the book:
Sebastien Bratieres,
	Kai Brodersen,
	      Frances Coll,
Daniel Galvez,
Abhishek Kumar,
Max Lepikhin,
Aaron Michelony,
	      Horst Stühler,
Hal Varian,
	      Clem Wang.

	      
      <li> People who have helped with the figures:
	Sandeep Choudhary, and others who are credited in the figure captions.
	      
		      
      <li> People who have helped with the code: Mahmoud Soliman,
	        Aleyna Kara,
  Srikar Jilugu,
  Drishti Patel,
  Ming Liang Ang,
  Gerardo Durán-Martín,
	      and other people listed 
	      <a href="https://github.com/probml/pyprobml/blob/master/README.md#acknowledgements">here</a>.
    </ul>