Optimizing Accuracy In Decision Making Using Evolutionary Computing

Analytic Hierarchy Process (AHP) is one of the methods in Decision Support Systems (DSS). AHP has been criticized mainly for its priority deviation method, which is one of AHP¿s main components. The priority derivation method, also referred to as prioritization method, is used to derive priorities in order to represent the rank of alternatives in AHP. There are two approaches to derive priorities in AHP, which are non-optimization approach and optimization approach. However, this study found three main problems in the current prioritization methods which are inconsistency of the judgment, non-evolutionary computing approach, and accuracy performance of the prioritization method. In solving these problems, this study proposed Evolutionary Computing Procedure for Deriving Priorities (ECPDP). The ECPDP is an EC-based procedure and derives priorities by solving single objective optimization problem (SOP) through maximizing the accuracy of the solution, by using Total Deviation (TD) as an objective function. The result by using ECPDP is more promising as opposed to the other prioritization methods in terms of TD value.