基于模糊距离和神经网络的自适应群决策方法

Adaptive group decision making method based on fuzzy distance and neural network

针对群决策中专家权重和评价指标权重难以确定的问题,提出了一种自适应群决策方法.设计了一种基于模糊距离的专家权重确定方法和基于神经网络的指标权重训练方法,并将其引入自适应算法中.在算法不收敛时,提出了一种专家权重的随机扰动策略使算法可以很好收敛.最后的实验表明,该方法可以很好地收敛于稳定解,得到专家群体对待评对象的客观、可靠评价.

An adaptive group decision making algorithm is proposed. A fuzzy distance method is proposed to establish the weights of experts and a neural network method to get the weights of the criteria is proposed. If the algorithm is not convergent, a strategy of stochastic perturbation will be used in order to get the convergent weights of experts and criteria. Experiments show that our methods can converge efficiently and get the objective and credible evaluation of the objects.