论文部分内容阅读
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。
Fuzzy method is an effective method of chemical pattern classification, but the acquisition of fuzzy rules and the determination of related parameters is more difficult. In this regard, this paper uses the rough set method, without any prior knowledge, reduce the system, access to the most simple rules set, on the basis of building a reasonable structure. Which is suitable for classification of fuzzy-neural network system, and initializes the network parameters according to the statistical properties of the rules and discretization results, and uses the LM method to train the network. In the application of the olive oil model classification modeling, the method has fast convergence training, Model prediction performance is good, better than the modern statistical methods and feedforward neural network.