Article ID: | iaor19981536 |
Country: | Netherlands |
Volume: | 86 |
Issue: | 2 |
Start Page Number: | 189 |
End Page Number: | 200 |
Publication Date: | Oct 1995 |
Journal: | European Journal of Operational Research |
Authors: | Smith Jim Q. |
Keywords: | graphical methods |
I argue here that the problem of coding uncertainty in environments with many sources of variation in terms of a probabilistic inferential system is now, for the most part, solved thanks to recent developments in influence diagrams. In this paper I outline why I believe this to be the case and review some of the more important features of this methodology which have enabled such complex probability systems to be elicited and to work efficiently.