| 摘要: | This study mainly cites the following works (and is an extension work to which): (1) Modelling the decision of paper shredder selection using analytic hierarchy process and graph theory and matrix approach (AHP–GTMA). Advances in Mechanical Engineering, 9(12), 1-11, 2017, and (2) Applying AHP+GTMA and AHP+TOPSIS on solving the same paper shredder selection problem and comparing the results. In: Proceedings of ICICE 2017. In (1), the recent AHP–GTMA approach had been applied as a decision aid in real practice to select the best paper shredder for a company as to make a bulk purchase order. In (2), another ‘AHP–TOPSIS’ selection decision method was introduced as an alternative multi-attribute decision-making (MADM) approach to solve the same problem, while the final results obtained by the two models (e.g., the rank orders) are observed and compared. However, as there is a critical question as to whether the use of one such recently proposed approach (AHP–GTMA) is effective to aid the selection decision problem in practice, the comparison process should be as rigorous as possible, but in (2) the comparison was made on a rough observational basis and this lacks of a solid basis that is weaved by several solid methods. In this study, as a remedy to this, a systematic flow to compare the main results from using these two methods (i.e., the rank order vectors) is proposed. When the similarity between the two obtained final rank orders is confirmed using the proposed method (to validate the observed similarity between the rank orders using a set of initial observable geometrical-based measures and then applying a statistical-based non-parametric test over the rank order vectors), the effectiveness of the AHP–GTMA approach is determined. It is shown the rank order among the alternatives that is determined by an AHP–GTMA is very similar to that which is obtained using an AHP–TOPSIS. Therefore, the effectiveness in solving a selection decision problem in practice using an AHP–GTMA is therefore verified by the verified similarity between the results. |
