Configurable DSS for Uncertainty Management by Fuzzy Sets

Abstract In this paper, we propose a Configurable Model Based DSS capable of dealing with generic problems being modeled by Linear Programming (LP) and by Fuzzy Sets (FS) in a deterministic and uncertain context, respectively. The DSS assumes the transformation of the original model with fuzzy coefficients into an equivalent crisp model where the fuzzy coefficients are represented as alpha-parametric values, which can vary in a predefined interval based on the alpha parameter. Through the DSS, solutions obtained by solving the deterministic model and the equivalent crisp model for different alpha-values are compared based on the objectives and performance parameters defined by the Decision Maker (DM). Due to the uncertainty in data, expected performance of solutions can change under real situations. The DSS allows simulating future real situations by generating different projections of uncertain parameters. New performance of previously generated solutions can be tested under these hypothetical real situations by means a third model (Model for the Real Performance Assessment). Finally, the DM can choose the solution to be implemented taking into account the performance of solutions under planned and real uncertainty.