粗糙集与模糊系统集成的化学模式分类方法及其应用

模糊方法是一种有效的化学模式分类方法,但模糊规则的获取和相关参数的确定较为困难。对此,本文采用粗糙集方法,无需任何先验知识,约简系统,获取最简规则集,在此基础上构建结构合理．适用于分类的模糊-神经网络系统,并根据规则的统计性质和离散化结果初始化网络参数,采用LM方法训练网络；在橄榄油模式分类建模的应用中,该方法训练收敛速度快,所建模型预测性能良好,要优于现代统计方法和前馈神经网络。

Integration of rough sets and fuzzy inference system for chemical pattern recognition and its application

The rough sets was introduced as the methods of identifying the structure of fuzzy system.First,the discretization of continu- ous attributes was made by the RSE-Chi2 methods,the redundancy of attributes and its value was eliminated by rough sets data analy- sis,and we got a collective rules about the chemical pattern classification system by identifying dependence in attributes from sample data.Second,a fuzzy-neuro networks system was built by these rules,the initial parameters' value of fuzzy-neuro networks was decided by the statistical parameters of rules and the result of discretization,and the networks was trained by LM methods.In problem of the or- igin discrimination of olive oil,the convergence speed of training is fast and the prediction of model is good when our methods was used,the results show the performance of our methods is superior to the methods of statistical and feedforward networks.