Kosana Suvočarev

Group Decision-Making in Selecting Nanotechnology Supplier

A group decision-making context is created to enable assessment of the Analytic hierarchy process (AHP) performance in presence of complete and incomplete information. To illustrate it, four recognized nanotechnology suppliers are evaluated across seven commonly used company/product attributes (net price, delivery, quality, production facilities, technical capability, management and organization, and geographical location) to identify the best one in multicriteria sense. Broad overview of applicable selection criteria and related MCDM methodologies is also presented.

Objective Evaluation of Walnut Cultivars by the Analytic Hierarchy Process

The walnut cultivars selection, supported by computerized multi criteria decision-making tools, has been initiated in Serbia 10 years ago. Recent case studies proved that Analytic Hierarchy Process (AHP) is a tool capable to integrate biological and technical aspects of the selection process and to link the expertise and standardization in fruit selection in general. AHP also performed very well in group contexts. In real-life AHP applications walnut experts liked to participate in computer-based decision-making sessions and commented that sessions were user-friendly and that derived weights, as global AHP outcomes indicating the walnut nuts’ quality, are of trustful reliability. However, analysts have been aware that experts’ opinions are based mostly on personal feelings about what they expected as the outcome of AHP session. To develop more objective approach, an original AHP prioritization method, eigenvector method (EV), which has been used in former applications to derive weights of compared decision elements, is compared with other two well known methods. Leading idea was to check which method produces ‘objectively trustful’ weights, i.e. assures best consistency and coherency, minimizes violation criteria, and all this - regardless of experts’ opinions. Two prioritization methods used along with eigenvector method are the logarithmic least square method (LLS) and fuzzy preference programming (FPP) method. Weights obtained by different decision makers are tested on consistency and it was shown that one of the two methods in some aspects outperform eigenvector method. Conclusion derived is that in real-life AHP applications an objective interpretation of results is necessity and that this can be achieved if several prioritization methods are used on competitive basis. Example includes 2 experts, 5 national and foreign cultivars and 7 criteria: kernel’s color, kernel’s portion, nut’s weight, taste of the kernel, shell, storage quality, and trade value.