1
$\begingroup$

I'm a student interested in learning about probabilistic networks — Bayesian Networks Markov models etc. I have a good background in probability, statistics and Markov chains.

Some good introductory books I found are the one by Jensen and Nielsen "Bayesian Networks and Decision Graphs", Springer and "Probabilistic Networks and Expert Systems" by Cowell et al., Springer.

Which one (of the above or not) would you recommend as a first read?

$\endgroup$
  • $\begingroup$ I have read an earlier version of the Jensen book and it is a good introduction: clear examples, but also going into the math and efficient algorithms. The Cowell book also great, but the math is a bit more difficult, and i found it less helpful as a starting guide. The main text i started with was Kjærulff's book. $\endgroup$ – user20650 Apr 16 '16 at 13:41
1
$\begingroup$

The bible on this is most likely: Probabilistic Graphical Models - Koller and Friedman. Here's the course website: http://online.stanford.edu/pgm-fa12

$\endgroup$
  • $\begingroup$ Just skimmed through it, looks great and complete. What is your opinion on the above two books as an introduction, though? $\endgroup$ – Rrjrjtlokrthjji Mar 22 '16 at 12:01
0
$\begingroup$

I suggest starting with Bayesian Networks and Influence Diagrams as it contains a lot of great examples. Koller's book is the bible on this but it is quite difficult to start with.

The Koller's book goes into great detail about all parts of PGMs including mathematical derivations. It contains everything one would need to know about the subject but it is very easy to get overwhelmed (it contains about 1200 pages but very few examples). The book that I mention does not go into so much details but gives you a very good overview of the subject and many worked out examples.

$\endgroup$
  • $\begingroup$ This is a bit short as an aswer, and looks more like a comment. Can you expand it? $\endgroup$ – Joe_74 Apr 28 '16 at 13:45
  • $\begingroup$ I agree with @GiuseppeBiondi-Zoccai , a summary of it's strong points in reference to the original questions would help to legitimize this as a full answer. $\endgroup$ – Underminer Apr 28 '16 at 13:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.