粗糙集理论框架下的神经网络建模研究及应用

为协调决策支持和分类,引入了一种新的方法,该方法将粗糙集理论和神经网络有机地结合在一起,提出了一种基于粗糙集理论的神经网络模型构造方法.首先,利用粗糙集理论智能数据分析的能力,对神经网络进行预处理,抽取关键成分作为神经网络的输入,从而确定粗糙神经网络的初始拓扑结构.在此基础上,进一步研究和分析了该模型的实现步骤,并应用原始数据对网络进行训练,最后将该模型应用于分类规则的抽取.试验结果比较表明,该模型可以有效地提高分类的精度.

On the Neural Network Modeling with Support Rough Set Theory

A method is introduced to cooperate the decision support and classification, in which rough set theory and neural network are formed integrated into a model. A neural network modelling way based on rough set theory is proposed. The neural network is preprocessed by the intelligent data analysis capability of rough set theory, and the key components are extracted as the inputs of the neural network to determine original topology of the rough neural network. Furthermore, the realization steps of the model are analyzed, and the rough neural network is trained with original data. The model constructed is applied to extraction of classification rules. The experimental results show that the model can increase the classification correctness effectively.