摘要
鉴于传统模糊神经网络计算繁琐、模型精度较低、难以收敛等问题,结合区间值与粗糙集理论,通过简化网络结构,改进输入条件,提出基于区间值推理的改进的模糊神经网络.通过仿真实验,验证了该方法的可行性.这一结果为研究模糊神经网络提供一种新的方法.
Considering the complicated computation, low model precision and convergence problem of the traditional fuzzy neural network, the interval value and the fuzzy reasoning are combined in this paper. By simplifying the network structure, and improving the input conditions and'the process of neural network construction, they are combined with the interval value of the rough set theory. Simulation experiments is carried out to verify the feasibility of ways. As a result, it presents a new method to study the fuzzy neural network.
出处
《沈阳理工大学学报》
CAS
2009年第4期39-43,共5页
Journal of Shenyang Ligong University
关键词
区间值
粗糙集
模糊神经网络
interval value
rough set
fuzzy neural network