Evaluation method based on interval AHP and DEA

Analytic Hierarchy Process (AHP) is a tool to assign importance grades to many items. However, a decision maker tends to give irrelevant importance grades because of intuitive judgement on pairwise comparisons. Thus a comparison matrix may include some inconsistent elements. Interval AHP for obtaining interval importance grades is proposed, where intervals reflect inconsistent elements in the given comparison matrix. Interval importance grades are obtained by interval regression analysis. To choose crisp importance grades for crisp efficiency, we use Data Envelopment Analysis (DEA), that is a tool to evaluate an object from the optimistic viewpoint considering many input and output items. The weight in DEA and the importance grade through AHP are similar to each other. We normalize data in order to make the weight obtained in DEA itself represent the importance grade in AHP.