Interactive construction of graphical decision models based on causal mechanisms
We propose a framework for building graphical decision models from individual causal mechanisms. Our approach is based on the work of Simon [Simon, H.A., 1953. Causal ordering and identifiability. In: Hood, W.C., Koopmans, T.C. (Eds.), Studies in Econometric Method. Cowles Commission for Research in Economics. Monograph No. 14. John Wiley and Sons Inc., New York, NY, pp. 49-74 (Ch. III)], who proposed a causal ordering algorithm for explicating causal asymmetries among variables in a self-contained set of structural equations. We extend the causal ordering algorithm to under-constrained sets of structural equations, common during the process of problem structuring. We demonstrate that the causal ordering explicated by our extension is an intermediate representation of a modeler's understanding of a problem and that the process of model construction consists of assembling mechanisms into self-contained causal models. We describe ImaGeNIe, an interactive modeling tool that supports mechanism-based model construction and demonstrate empirically that it can effectively assist users in constructing graphical decision models.
|Year of publication:||
|Authors:||Lu, Tsai-Ching ; Druzdzel, Marek J.|
European Journal of Operational Research. - Elsevier, ISSN 0377-2217. - Vol. 199.2009, 3, p. 873-882
|Keywords:||Decision analysis Model building Bayesian networks Structural equation models Causal ordering|
|Type of publication:||Article|
Persistent link: https://www.econbiz.de/10004973602
Saved in favorites
Similar items by person
Lu, Tsai-ching, (2009)
Lu, Tsai-Ching, (2009)
- More ...