The kernel-based comprehensive aggregation PROMETHEE (PROMETHEE-KerCA) method for multi-criteria decision making with application to policy modelling

Abstract

As the economic and technological problems become more complex and require effective multi-criteria decision making (MCDM) tools for analysis thereof, there is a need for comprehensive MCDM techniques that would be capable to ensure robust optimization with minimum arbitrary assumptions. This paper proposes a new method for MCDM – the Kernel-based Comprehensive Aggregation PROMETHEE (PROMETHEE-KerCA). The proposed approach relies on the kernel density estimation which provides the bandwidths for scaling the differences in the performance of the alternatives. The kernel-based distances are aggregated to establish the performance measures thus following the principle of the outranking. Then, the measures of performance are aggregated in four different manners (additive, multiplicative, minimum and maximum values) to construct the comprehensive overall utility score. The proposed method does not require choosing the preference functions or parameters thereof. The empirical illustration is provided to show the feasibility of the proposed approach. The European Union Member States are ranked by the means of the KerCA method with regards to the objectives of the strategy Europe 2020. The isolated and pooled ranking allows comparing the progress of the countries compared with their initial situation and compared to the other countries in the sample.