A decision support tool for the assessment of alternatives using fuzzy set theory and classical methods

Abstract

Current solution procedures for multi-attribute computer aided support tools tend to neglect adequate assessment of subjective aspects which cannot be measured precisely in an objective manner. Thus, qualitative types of imprecision like inexactness and vagueness are not taken into account. The combined use of fuzzy set theory with classical assessment methods has not received much attention either. The objective of this thesis was to develop a decision support tool to assess multi-attribute alternatives using fuzzy sci theory and extending classical methods. Algorithms developed include: modifying the classical midvalue splitting method, and incorporating confidence ratings when assessing scaling constants. Jain's weighted rating method was used to calculate the final results. The FUZZY package is developed based on these algorithms, and its validity was tested by an experimental case study.