Business Rules Technology and Rule-Based Systems

"Inference Engines for Everyone"

Aut inveniam
viam aut faciam
Home News About This Site All Pages All Tags Wiki


Updated Sept 16 2008

An interesting collection of people and papers at the UCLA Graduate Summer School 2007, Probabilistic Models of Cognition: The Mathematics of Mind ( encrypted for some reason ). What better way to spend one's summer.

A notable presentation is by Robert Jacobs on Bayesian Decision Theory ( PDF ). He starts out with 'signal detection theory' and gives an example of using the Bayesian approach to determine whether a given object is an apple or an orange. At what mathematically exact point of lightness ( or redness or whatever ) should one decide whether the object is an apple or an orange ?

This simple example introduces the difficult topic of how vague and fuzzy human perceptions about the world become concrete decisions to act or not to act. For example, how do perceptions about a continuous stream of numbers such as the Dow Jones Industrial Average become discrete decisions to buy or sell a certain stock ? The necessity for binary decisions in a non-binary world produces sharp and somewhat artificial decision points that cause much of the 'brittleness' associated with rule-based systems.

Warning: there's some math in this presentation, but not too much. It is a complex subject and Robert Jacobs makes it about as accessible as it gets.

A decent Subject Hierarchy for Data and Knowledge ( PDF ) at Elsevier. It was written in 2002 and may be a bit outdated.



Tags:   Knowledge   Engineering

Other BBcom-related sites - Quick Links




A  Dial-Up Friendly Site  

We Do SVGA ( mostly )

Hisssss ...