Influence Diagrams under Partial Information
Concha Bielza, D R íos-Insua, S R íos-Insua · 1996
Abstract We consider influence diagrams in which only partial information is available about probabilities and utilities. Algorithms for computation of optimal alternatives in ID’s may be generalised to compute non-dominated alternatives in ID’s under partial information. Typically, the set of nondominated alternatives is too big so we introduce additional criteria to select one of those alternatives.