一种利用神经网络改善判断矩阵一致性的方法

摘要

With respect to the problem of improving the consistency of judgement matrix in analytic hierarchy Process, a method is proposed to improve the consistency of judgement matrix by using neural network.Based on the BP neural network model, this paper transforms the problem of adjusting the consistency of judgement matrix into the problem of multi-input and multi-output solution of BP neural network.The judgement matrix return a-gain for the expert further is adjusted after the BP neural network algorithm is adjusted to conform to Saaty scale.Numerical examples indicate that this method is effective in practice.%针对层次分析法中判断矩阵致性改进问题,提出了一种利用神经网络改善判断矩阵一致性的方法.本文在建立了BP神经网络模型的基础上,把判断矩阵一致性调整问题转化为BP神经网络的多输入多输出求解问题.经BP神经网络算法调整过的判断矩阵再返回给专家进一步调整使其符合萨迪标度.计算实例表明,此种方法是可行的.