【摘要】 文章介绍了一种基于进化式模糊神经网络时间预测系统,它是一种快速自适应的局部学习模型;进化式模糊神经网络是一个特殊类型的神经网络,它能通过进化其结构和参数来容纳新的数据。文章重点介绍了网络结构、学习方法及创建、修剪、聚合规则节点的算法;实验结果表明:模糊隶属函数的个数,规则的修剪和聚合等训练参数,与网络的行为和预测结果有很重要的关系。更多还原

【Abstract】 Introduces a time series prediction system basing on evolving fuzzy neural network.It is a local learning model which allows for fast adaptive learning.Evolving fuzzy neural network is a particular type neural network,which evolves both its structure and parameters to accommodate new coming data.It is an important introduction on network structure,learning method and algorithm of nodes creating,pruning,aggregation.Experiment results demonstrate that there are important relations between the parameters on numbers of fuzzy member function,rules pruning and aggregation etc. and network behavior,prediction results.更多还原