New approach to multiple attribute decision making with interval numbers

• 作 者: Zhang Quan Gao Qisheng Geng Ji
• 机 构: School of Information Engineering, Shenyang Univ. Of Technology, Shenyang 110023, P. R. China;Inst.
• 出 处: 系统工程与电子技术（英文版）
• 发表日期: 2008年2期
• 关键词: multiple attribute decision making interval number normalization programming mod

In an ambiguous decision domain, the evaluation values of altatives against attributes would be interval numbers because of the inherent, uncertain property of the problems. By using a number of linear program-ming models, Bryson and Mobolurin propose an approach to compute attribute weights and overall values of the altatives in the form of interval numbers. The intervals of the overall values of altatives are then transformed into points or crisp values for comparisons among the altatives. However, the attribute weights are different because of the use of linear programming models in Bryson and Mobolurin’s approach. Thus, the altatives are not comparable because different attribute weights are employed to calculate the overall values of the altatives.A new approach is proposed to overcome the drawbacks of Bryson and Mobolurin’s approach. By transforming the decision matrix with intervals into the one with crisp values, a new linear programming model is proposed, to calculate the attribute weights for conducting altative ranking.