Prognostic and Diagnostic Capabilities of OOBN in Assessing Investment Risk of Complex Construction Projects

Abstract

Modelling decision problems using Bayesian networks is extremely valuable especially in case of issues related to uncertainty; it is also very helpful in constructing and understanding visual representation of the elements and their relations. This approach facilitates subsequent application of Bayesian networks, however there can be situations where using simple Bayesian networks is impractical or even ineffective. The aim of this article is to present object-oriented Bayesian networks (OOBN) in the context of modeling investment risk. OOBN not only allow decomposition of a complex model into individual objects reflecting different groups of issues (for example risk areas) but also allow modeling time dependencies between those objects.The use of object-oriented Bayesian networks is presented using an example of urban regeneration project. On the basis of a complex construction project the author presents both advantages and disadvantages of OOBN in terms of diagnostic and prognostic efficiency.In course of the research it has been observed that during the construction of large Bayesian networks the possibility to automatically generate node probability tables is very useful, as it significantly accelerates construction of this type of models. The author also indicates additional recommendations in the field of defining object-oriented Bayesian networks instrumental in assessing investment risk of complex construction projects.