Coenen / Allen Research and Development in Intelligent Systems XXII

Proceedingas of AI-2005, the Twenty-fifth SGAI International Conference on Innovative Techniques and Applications of Artificial Intelligence
1. Auflage 2010
ISBN: 978-1-84628-226-3
Verlag: Springer
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, 358 Seiten

The papers in this volume are the refereed technical papers presented at AI2005, the Twenty-fiftth SGAI International Conference on theory, practical and application of Artificial Intelligence, held in Cambridge in December 2005. The papers in this volume present new and innovative developments in the field, divided into sections on Machine Learning, Knowledge Representation and Reasoning, Knowledge Acquisition, Constraint Satisfaction and Scheduling, and Natural Language Processing. This is the twenty-first volume in the Research and Development series. The series is essential reading for those who wish to keep up to date with developments in this important field. The Application Stream papers are published as a companion volume under the title Applications and Innovations in Intelligent Systems XIII.

Autoren/Hrsg.

Coenen, Frans

Allen, Tony

Leseproben

"SESSION 2: NETWORKS AND BIOLOGICALLV MOTIVATED AI (S. 132-133)

Exploring the Noisy Threshold Function in Designing Bayesian Networks*

Rasa Jurgelenaite, Peter Lucas and Tom Heskes Radboud University Nijmegen, Nijmegen, The Netherlands E-mail : {rasa.peterl.tomh}@cs.ru.nl

Abstract
Causal independence modelling is a well-known method both for reducing the size of probability tables and for explaining the underlying mechanisms in Bayesian networks. Many Bayesian network models incorporate causal independence assumptions; however, only the noisy OR and noisy AND, two examples of causal independence models, are used in practice. Their underlying assumption that either at least one cause, or all causes together, give rise to an effect, however, seems unnecessarily restrictive. In the present paper a new, more flexible, causal independence model is proposed, based on the Boolean threshold function. A connection is established between conditional probability distributions based on the noisy threshold model and Poisson binomial distributions, and the basic properties of this probability distribution are studied in some depth. The successful application of the noisy threshold model in the refinement of a Bayesian network for the diagnosis and treatment of ventilator-associated pneumonia demonstrates the practical value of the presented theory.

1 Introduction

Bayesian networks offer an appealing language for building models of domains with inherent uncertainty. However, the assessment of a probability distribution in Bayesian networks is a challenging task, even if its topology is sparse. This task becomes even more complex if the model has to integrate expert knowledge. While learning algorithms can be forced to take into account an experts view, for the best possible results the experts must be willing to reconsider their ideas in light of the models discovered structure.

This requires a clear understanding of the model by the domain expert. Causal independence models can both limit the number of conditional probabilities to be assessed and provide the ability for models to be understood by domain experts in the field. The concept of causal independence refers to a situation where multiple causes independently influence a common effect. Many actual Bayesian network models use causal independence assumptions . However, only the logical OR and AND operators are used in practice in defining the interaction among causes; their underlying assumption is that the presence of either at least one cause or all causes at the same time give"