Learning in Graphical Models (Adaptive Computation and Machine Learning)

Graphical models, a marriage between probability theory and graph theory, provide a natural tool for dealing with two problems that occur throughout applied mathematics and engineering--uncertainty and complexity. In particular, they play an increasingly important role in the design and analysis of machine learning algorithms. Fundamental to the idea of a graphical model is the notion of modularity: a complex system is built by combining simpler parts. Probability theory serves as the glue whereby the parts are combined, ensuring that the system as a whole is consistent and providing ways to interface models to data. Graph theory provides both an intuitively appealing interface by which humans can model highly interacting sets of variables and a data structure that lends itself naturally to the design of efficient general-purpose algorithms.

This book presents an in-depth exploration of issues related to learning within the graphical model formalism. Four chapters are tutorial chapters--Robert Cowell on Inference for Bayesian Networks, David MacKay on Monte Carlo Methods, Michael I. Jordan et al. on Variational Methods, and David Heckerman on Learning with Bayesian Networks. The remaining chapters cover a wide range of topics of current research interest.

To its credit, "Learning in Graphical Models" has several well-written and interesting papers, but the tutorial papers just did not seem enough of an introduction for me to feel comfortable using it as a first source of introduction.

What I find most compelling about Bayesian networks is the fact that they seem both highly modular (which facilitates reusability and network interconnectivity) and can be designed in a semi-rational manner (contrast this with neural-network architectures for which few good algorithms exist for determining size and number of layers). For this reason I imagine they will be important players in future engineering projects that require learning and adaptation.


Simply Superb...   March 31, 2000
11 out of 28 found this review helpful

My area of research revolves around graphical models... Best Book... The book that introduced me as to how effective graphical models are... As stated in the editorial review, graphical model is the marriage between graph theory and probability and this book justifies the sacredness of this marriage!