Graphical Models and Inference - Lecture 8

Bayesian Networks and Expert Systems

This lecture describes the general idea behind expert systems and the use of Markov models based on DAGs, known as Bayesian networks. The methods are illustrated with examples from forensic DNA analysis, in particular paternity cases.

General reading material for this section would be the first 1-3 chapters of Cowell et al. (1999). Further details on the application of Bayesian networks in DNA analysis can for example be found in

Dawid, Mortera, Pascali and van Boxel (2002). Probabilistic Expert Systems for Forensic Inference from Genetic Markers, Scandinavian Journal of Statistics, 29 577-595.

Overheads

Links

Previous lecture

Course overview


Last updated: Friday, 28 October 2011Steffen L. Lauritzen