Probabilistic Graphical Models: Principles and Techniques (Adaptive Computation and Machine Learning series)

great book. must read for machine learning and a good reference book

If you want a very close look under the hood of Bayesian Networks, I can highly recommend Probabilistic Graphical Models. It's extremely comprehensive (1,200+ pages), well structured and clearly written. Theory, computation and application - including how to think about causation - are all covered in depth. Not light reading and not suited for those with limited stats background, but all in all one of the best textbooks on analytics topics I've ever read. Very impressive.

This is the textbook for my PGM class. It is definitely not an easy book to read, but its content is very comprehensive. It is a great reference to get more details of PGM. I highly recommend this book!

I have read this book in bits and pieces and find it extremely useful. Finally, we got a book that can be used in classroom settings. There are some typos (hence four stars) that will hopefully get fixed in the future editions. The book also has a lot of new insights to offer that can only be gleaned from the vast existing literature on the topic with excruciating labor. Agreed that this book is pricey but for what it has to offer, I think it was money well spent.

This is a great book on the topic, regardless of whether you are new to probabilistic graphical models or have some familiarity with them but would like a deeper exploration of theory and/or implementation. I have read a number of books and papers on this topic (including Barber's and Bishop's) and I much prefer this one. Dr. Koller's style of writing is to start with simple theory and examples and walk the reader up to the full theory, while adding reminders of relevant topics covered elsewhere. She accomplishes this without condescending to or belittling the reader, or being overly verbose; each of the 1200 pages is concise and well edited. There is an OpenClassroom course that accompanies the book (CS 228), which I highly recommend viewing, as it contains that same style of teaching but in a different format and often with a somewhat different approach.

I have learned basics of graphical models from a professor who is quite prominent in the field. He taught it from unpublished book by Michael Jordan + few chapters by Chris Bishop. I have not read most of the books but have read enough to write positive things about it. I especially like the part of the book that shows dependencies (bad pun alert). dependencies of chapters that is. :D the only complaint i have is not towards the authors but towards the publishers. the quality of paper is the worst i've ever seen and i own more than 400 textbooks. there are dusts all over the pages. you can feel your hands getting dry due to these paper particles and after a while you can't breathe because of these particles. some books have this but this book is the worst when it comes to that paper dust. you will know when you have this yourself. they could have slapped on $200 and worse paper quality, I would still buy it without thinking twice about it.

This popular book makes a noble attempt at unifying the many different types of probabilistic models used in artificial intelligence. It seems like a good reference manual for people who are already familiar with the fundamental concepts of commonly used probabilistic graphical models. However, it contains a lot of rambling and jumping between concepts that will quickly confuse a reader who is not already familiar with the subject. While the book appears to be systematic in introducing the subject with mathematical rigor (definitions and theorems), it actually skips a lot of fundamental concepts and leaves a lot of important proofs as exercises. I would recommend that a beginner in the subject start with another book like that by Jordan and Bishop, while keeping this book around as a reference manual or bank of practice problems for further study. The Coursera class on this subject is much easier to follow than this book is.

with an eye to taking the course. Very informative. Although the phrase "in context" covers a multitude of sins. I'd prefer the distinction between the the distribution of an intersection of random variables (where comma's are used as a short-hand) and joint distributions a bit clearer. Aside, I managed to find an error not listed on the errata web page for the book. The equation for MAP queries on page 26 has it as the maximal assignment of a JOINT distribution, while on the next page it is the maximal assignment of a CONDITIONAL distribution (I believe this is the correct one). This was a little confusing until I read page 26 a bit closer. Before you ask, yes I do read Math textbooks for pleasure.

It's a great, authoritative book on the topic - no complains there. My one issue is that the shipped book is not colour but gray-scale print. I was hoping that's the least I could expect after paying over $100 on a book. Graphs and charts are imperative to reading technical books such as this, and anyone remotely familiar with ML/Statistics will agree with me that having coloured charts make an immense difference in this field. Is this book shipped from the actual publishers or a 3rd party vendor? Also, it'll be helpful to explicitly state the gray-scale print in description to manage expectations. Thanks.

Un testo notissimo e completo sull'argomento, senz'altro complesso ed esaustivo. Adatto all'autoapprendimento a partire da un buon livello di conoscenza.

PGMs sind 'die' Erfolgsstory der KI. Einer der bekanntesten Protagonisten ist J.Pearl, der Preisträger des Turing Award 2011. Pearl "erfand" vor 1988 eine spezielle Variante der PGM, die als Bayesian Belief Networks (BBN) bekannt wurden. Die Einsatzgebiete reichen vom Bau von Assistenzsystemen über Bildverarbeitung bis hin zu Anwendungen in der Genetik und Robotik ("Google-car" von Sebastian Thrun). Es gibt eine Menge guter Bücher über PGM und BBN, darunter auch zwei, die von Pearl selbst geschrieben wurden. Jedoch die umfassendste und klarste Darstellung der Domäne gelang anhand vieler Beispiele Daphne Koller auf 1233 Seiten. Es werden im Teil I gerichtete (BBN) und ungerichtete PGM (Markov Nets) und der Modellbau mit Schablonen (Template Models) präsentiert. Teil II behandelt exakte und approximative Inferenzmethoden. Teil III ist dem Lernen der Parameter und der Modellstruktur gewidmet. Teil IV hat den Titel "Actions and Decisions". Dahinter verbergen sich Kapitel über die Repräsentation von Kausalitäten mittels sogenannter "Twin Networks" sowie Einflussstrukturen mittels "Influence Diagrams". Ich will jetzt hier nicht im Einzelnen auf die Inhalte eingehen. Aber soviel kann man sagen, dass diese Inhalte viele andere modische Inhalte der KI wohltuend überdauert haben und in anderen Wissenschaften (von Kognitionswissenschaft bis hin zur Elektrotechnik) rezipiert werden. Wer nicht die Zeit hat, das Buch von A-Z durchzuarbeiten, der sollte sich die gleichnamige kostenlose eLearning-Vorlesung von Daphne Koller auf coursera.org ansehen und durcharbeiten. In 11 Wochen werden die wichtigsten Inhalte des Buches durchgeackert. Dabei ist es günstig, das Buch als begleitendes Nachschlagewerk neben dem Lernrechner zur Verfügung zu haben. Kurzum PGM von Daphne Koller ist ein sehr empfehlenswertes Standardwerk, das sich aber eher um die mathematischen Grundlagen der PGM und weniger um den anwendungsbezogenen Modellbau kümmert. Die Beispiele stammen alle aus Bereichen, in denen Echtzeitrestriktionen keine Rolle spielen. Wer so etwas sucht, sollte zum Buch von Thrun (Probabilistic Robotics) greifen. Um Platz zu sparen, gibt es bei Behandlung der Beispiele viele Verweise auf vorhergehende Seiten. Das erfordert häufiges nerviges Hin- und Herblättern, was den sehr guten Gesamteindruck aber nur unwesentlich mindert.

This is the book of the (free MOOC) course "Probabilistic Graphical Models" of Coursera. Really good purchase. Totally recommended if you are interested in advanced statistics.

It's a comprehensive book covering a diverse number of topics in probabilistic graphical models. The initial chapters discuss fundamental concepts which serves as background for the advanced chapters