Deriving priority in AHP using Evolutionary Computing approach

Abstract

In the real world, human will face the problem and dilemma to making decision. Making decision is the critical part in choosing the best solution. Multi-criteria decision making (MCDM) is one of the most well known branches of decision making and it is referring to making decision in the presence of multiple criteria. MCDM problem are common occurrences in everyday life. In 1977, Saaty introduced Analytic Hierarchy Process (AHP) to solve the MCDM problem. The AHP is widely used for MCDM. Since AHP has been introduced, it has been applied in numerous situations with impressive results. However, AHP has been also criticized in the literature review, mainly in priority derivation procedure. This paper has identified three main problems in current priority derivation procedure which are: (1) Inconsistency of the judgment, (2) Non-evolutionary computing approach, and (3) Accuracy performance of the prioritization method. To solve the criticism and the problems; this paper proposes AHPEC which is using Evolutionary Computing (EC) to derive priorities in AHP. The AHPEC gives better result compare to the other prioritization methods based on accuracy of derived priorities. The comparison is based on the value of Total Deviation (TD) which is measure accuracy o the solution. The case study from Srdjevic, 2005 was chosen to compare the performance of the AHPEC and the current prioritization methods based on accuracy of the solution as a criterion to be optimized.