Graphical Models refers to a unified approach to understanding and developing algorithms across a variety of methodological domains including Bayesian Nets, Neural Nets, A.I., probabilistic reasoning, inference, learning, prediction, feature extraction, basis reduction (ICA, PCA, Factor Analysis, Projection Pursuit), regression, HMM's, etc.
This site is intended to help quickly orient newcomers to the GM perspective.

[If you are looking for info about our GM Reading Group at NUS, please visit here.]

Good Introductions to Graphical Models
    0) The Introduction to the book Graphical Models; Foundations of Neural Computing (ed. M. Jordan) - is a great place to gain an initial understanding of what a GM is. If you find it on-line, let me know!
    1) R.G. Cowell (1998), Introduction to inference in Bayesian networks. in Learning in Graphical Models (ed. M. Jordan) - is foundational, technical and succinct.
    2) Smyth, Heckerman & Jordan, Probabilistic Independence Networks for Hidden Markov Probability Models. in Graphical Models; Foundations of Neural Computing (ed. M. Jordan) - like Cowell (above) covers the basics of constructing junction trees for exact inference, but the latter part of the chapter focuses on generalizing HMMs with GM techniques. Gives a good sense of the power of the GM paradigm.
    3) David Heckerman's Microsoft technical report, A Tutorial on Learning With Bayesian Networks is a little broader in the learning algorithms it reviews than the others mentioned here.
    4)  Jordan, M.I, Ghahramani, Z., Jaakkola, T.S., and Saul, L.K. (1999)  - "An Introduction  to Variational Methods for Graphical Models" is a very accessible presentation of this approximate inferencing approach applied to several different classic model types including HMMs and feedforward NNs. This authoring team deserves an award for the clarity of their presentations.

    Keep an eye out for Michael Jordan's forthcoming book, an introduction to graphical models.

    Other introductory material on line:
    5) One-(long) page intro by Kevin Murphy

Graphical Model Code Resources
    Kevin Murphy's MATLAB Toolbox
    Kevin Murphy's list of related available software

Graphical Model Links
    Bayesian Network Repository   Standard datasets, the XML "Bayesian Interchange Format" description, other links.
    Many of the papers that appear in Learning in Graphical Models(ed. M. Jordan) are gathered.
    A USC course with great intro and background links (including some draft chapters of M. Jordans as-yet-upnpublished book).
 

    Saul and Jordan Neural Network - A JAVA implementation of the net described in  "Attractor dynamics in feedforward neural networks",  Neural Computation 12:6, 1313-1335, and in Chapter 5 in the book "Graphical Models; Foundation in Neural Computing", Michael Jordan and Terrence Sejnowski (Eds), MIT Press, 2001. Code submitted by Lonce Wyse.

Background Knowledge

Basic Probability Theory
Linear Algebra

The best place to find clear explanations of these things is in undergraduate textbooks.
A on-line resource for math basics is:  http://mathworld.wolfram.com/letters/ (see the subject index)
Matrix Algebra useful for statistics (Thomas Minka) - also includes MATLAB code
MATLAB is excellent for exploring, plotting, seeing how things work.

Related "Foundations" Reading:

Why the logistic function? A tutorial discussion on probabilities and neural networks. M. I. Jordan.
Glossary of Statistical learning terms and concepts from Thom Minka.
A Gentle Tutorial on the EM Algorithm from Jeff Bilmes.

Other Link Pages
Felix Agakov links for probability, AI, and GMs