BAYESIAN AND NEURAL NETWORKS FOR GEOGRAPHIC INFORMATION-PROCESSING

Authors
STASSOPOULOU A PETROU M KITTLER J
Citation
A. Stassopoulou et al., BAYESIAN AND NEURAL NETWORKS FOR GEOGRAPHIC INFORMATION-PROCESSING, Pattern recognition letters, 17(13), 1996, pp. 1325-1330
Citations number
11
Categorie Soggetti
Computer Sciences, Special Topics","Computer Science Artificial Intelligence
Journal title
Pattern recognition letters → ACNP
ISSN journal
01678655
Volume
17
Issue
13
Year of publication
1996
Pages
1325 - 1330
Database
ISI
SICI code
0167-8655(1996)17:13<1325:BANNFG>2.0.ZU;2-S
Abstract
In this paper we study the problem of obtaining a correspondence betwe en Bayesian networks and neural networks. It is shown how such a corre spondence is established by obtaining a mathematical function which re lates the parameters of the two models. We show the validity of our me thod by deriving the parameters to be used in a Bayesian network const ructed to combine GIS data for assessing the risk of desertification o f burned forest areas in the Mediterranean region.