基于人工神经网络的多属性决策

研究利用ＢＰ人工神经网络解决多属性决策问题。根据价值函数的存在定扩展多属性决策问题的价值函数，利用一个３层ＢＰ网络任意逼迫多属性决策问题的价值函数，从问题本身抽取学习样本训练ＢＰ网络，构造该问题的价值函数，利用训练后的ＢＰ神经网络作为多属性决策问题的价值函数计算各方案的价值。实现了多属性决策问题的自动化，计算结果与用传统工具的计算结果基本一致，并且满足Ｐａｒｅｔｏ最优准则。

Neural Network Based Multiple Criteria Decision Making

To solve the multiple criteria decision making problems, the multiattribute value function was extended based on the existence of the value function. A 3 layered BP neural network was used to construct the value function from the samples extracted from the multiple criteria decision making problems, then the calculated values of every plan were compared. The automation of multiple criteria decision making is implemented, the results agree with the results of the traditional algorithms and meet the Pareto optimal criterion.