基于粗糙集理论的多属性评估方法的研究

Multi-attribute evaluation method based on rough set theory

传统的模糊综合评价方法利用模糊数学理论来描述评价过程中涉及的一些模糊因素,取得了一定效果。但这种方法存在的隶属度参数调整的主观性、网络结构不易确定、模糊规则的选取等问题不易解决。本文提出结合粗糙集理论(RS)和模糊神经网络(FNN)的一种网络评估模型。使得网络的结构容易确定,并利用Rosetta软件轻松提取模糊规则。是一种具有通用性的网络评估模型。

The original fuzzy comprehensive evaluation uses fuzzy math theory to describe the process of evaluation and some fuzzy factors. This evaluation model is feasible. Some problems consist in the model ,such as how to adjust the parameters of membership functions,how to design the structure of the model and how to select the fuzzy rules. This paper designs a new model that combines rough set （RS） and fuzzy neural network（FNN）. It is easy to design the structure of the model and pick up the fuzzy rules with Rosetta. It is a evaluation model that can used in many cases.