Path: utzoo!utgpu!watmath!maytag!watdcsu!lhohner
From: lhohner@watdcsu.waterloo.edu (L.Hohner - Statistics)
Newsgroups: comp.ai
Subject: uw.stats.tutorial: An Introduction to Statistical Methods in Expert Sys
Keywords: rule-based systems, conditional independence, Markov models
Message-ID: <5670@watdcsu.waterloo.edu>
Date: 14 Mar 89 16:27:24 GMT
Distribution: ont
Organization: U of Waterloo, Ontario
Lines: 23


T U T O R I A L
DEPARTMENT OF STATISTICS AND ACTUARIAL SCIENCE
UNIVERSITY OF WATERLOO
SPEAKER:      Dr. David Spiegelhalter
              Medical Research Council
              Cambridge University
TITLE:        An Introduction to Statistical Methods
              in Expert Systems
ABSTRACT
An overview of the probabilistic representation of uncertainty in expert
systems will be provided.  Topics will include the inadequacy of rule-based
systems, the use of acyclic directed graphs to express conditional
independence, directed Markov models, computational algorithms based on
local message-passing, Markov field representations, relation between a
triangulated graph and a clique-tree, simple message-passing solution,
updating conditional probabilities from data, eliciting initial models,
and so on.
The talk will be reasonably non-technical, and will concentrate on
imparting a feel for the area with the aid of practical problems.
DATE:                   Friday, March 17, 1989
TIME:                          10:30 a.m.
PLACE:                          MC 5158