Methods for decision making with cardinal numbers and additive aggregation

For multi-attribute decision making with cardinal numbers and additive aggregation the methods ''additive aggregation'' and ''outranking'' for crisp and fuzzy numbers are connected with the concept of linguistic variables. Preference ranking a generalization of outranking and preference ranking organization methods - such as ELECTRE and PROMETHEE, can be divided into three phases, where each of them is defined completely by linguistic variables. The derived method ''preference ranking with additive weighting and individual fuzziness'' has an improved selectivity or discrimination compared to conventional methods. Moreover the method fulfils four important requirements: preference only for sufficiently large weights, considering evaluation differences outside of thresholds and regarding aggregational and substitutional properties. The simplified method ''preference ranking with additive weighting and general fuzziness'' allows a reduction of computational resources. (C) 1997 Elsevier Science B.V.