valisevskis - Ph.D. Student, Department of Modelling and Simulation, Riga Technical University

In this paper an overview of conditional-fuzzy-granule-based decision aid methodology is presented. We describe the main steps that this methodology consists of and give an overview of methods that are employed therein. This methodology concerns decision analysis under various sources of uncertainty, as both randomness and fuzziness is adapted in this methodology. Thus, it can be used when the information available is uncertain and incomplete. Another important point about this methodology is that ranking of alternatives is based on the probabilities that criteria will take desired values, which eliminates the scaling problem. The main aspect of the methodology is that the problem statement may be very uncertain, up to the point when the information available may not be enough to construct a performance table. That is the main difference between this methodology and more common ones. Description of the methodology is followed by a case study, which considers a project selection problem.

project selection, decision analysis under uncertainty, fuzzy logic, f-granules