Statistical Science

Bayesian Analysis in Expert Systems

David J. Spiegelhalter, A. Philip Dawid, Steffen L. Lauritzen, and Robert G. Cowell

Full-text: Open access

Abstract

We review recent developments in applying Bayesian probabilistic and statistical ideas to expert systems. Using a real, moderately complex, medical example we illustrate how qualitative and quantitative knowledge can be represented within a directed graphical model, generally known as a belief network in this context. Exact probabilistic inference on individual cases is possible using a general propagation procedure. When data on a series of cases are available, Bayesian statistical techniques can be used for updating the original subjective quantitative inputs, and we present a set of diagnostics for identifying conflicts between the data and the prior specification. A model comparison procedure is explored, and a number of links made with mainstream statistical methods. Details are given on the use of Dirichlet prior distributions for learning about parameters and the process of transforming the original graphical model to a junction tree as the basis for efficient computation.

Article information

Source
Statist. Sci. Volume 8, Number 3 (1993), 219-247.

Dates
First available in Project Euclid: 19 April 2007

Permanent link to this document
https://projecteuclid.org/euclid.ss/1177010888

Digital Object Identifier
doi:10.1214/ss/1177010888

Mathematical Reviews number (MathSciNet)
MR1243594

Zentralblatt MATH identifier
0955.62523

JSTOR
links.jstor.org

Keywords
Graphical models subjective probability conditional independence local computation triangulation junction tree unsupervised learning Dirichlet distribution Bayes factors prequential analysis prediction monitors

Citation

Spiegelhalter, David J.; Dawid, A. Philip; Lauritzen, Steffen L.; Cowell, Robert G. Bayesian Analysis in Expert Systems. Statist. Sci. 8 (1993), no. 3, 219--247. doi:10.1214/ss/1177010888. https://projecteuclid.org/euclid.ss/1177010888


Export citation

See also

  • See Comment: A. P. Dempster. [Bayesian Analysis in Expert Systems]: Comment: Assessing the Science Behind Graphical Modelling Techniques. Statist. Sci., Volume 8, Number 3 (1993), 247--250.
  • See Comment: Clark Glymour, Peter Spirtes. [Bayesian Analysis in Expert Systems]: Comment: Conditional Independence and Causal Inference. Statist. Sci., Volume 8, Number 3 (1993), 250--257.
  • See Comment: Joe R. Hill. [Bayesian Analysis in Expert Systems]: Comment. Statist. Sci., Volume 8, Number 3 (1993), 258--261.
  • See Comment: David Madigan. [Bayesian Analysis in Expert Systems]: Comment: What's Next?. Statist. Sci., Volume 8, Number 3 (1993), 261--263.
  • See Comment: Sharon-Lise Normand. [Bayesian Analysis in Expert Systems]: Comment. Statist. Sci., Volume 8, Number 3 (1993), 263--265.
  • See Comment: Judea Pearl. [Bayesian Analysis in Expert Systems]: Comment: Graphical Models, Causality and Intervention. Statist. Sci., Volume 8, Number 3 (1993), 266--269.
  • See Comment: Michael E. Sobel. [Bayesian Analysis in Expert Systems]: Comment. Statist. Sci., Volume 8, Number 3 (1993), 269--273.
  • See Comment: Joe Whittaker. [Bayesian Analysis in Expert Systems]: Comment. Statist. Sci., Volume 8, Number 3 (1993), 273--276.
  • See Comment: D. R. Cox, Nanny Wermuth. [Bayesian Analysis in Expert Systems]: Rejoinder. Statist. Sci., Volume 8, Number 3 (1993), 276--277.
  • See Comment: David J. Spiegelhalter, A. Philip Dawid, Steffen L. Lauritzen, Robert G. Cowell. [Bayesian Analysis in Expert Systems]: Rejoinder. Statist. Sci., Volume 8, Number 3 (1993), 277--283.