单体模糊神经网络的学习规则及其收敛性研究

RESEARCH ON THE LEARNING RULES AND THEIR CONVERGENCE OF MONOLITHIC FUZZY NEURAL NETWORKS

摘要: 梁久祯教授在不久前研究了单体模糊神经网络 (MFNNs)的函数逼近能力 .在此基础上 ,提出了单体模糊神经网络 (MFNNs)的学习规则并进一步研究了其收敛性 .研究结果表明 ,所提出的学习规则是收敛的 ,这一结论为单体模糊神经网络的应用提供了坚实的理论基础 .

Abstract: J Z Liang presents monolithic fuzzy neural networks （MFNNs） and studies their function approximation capabilities. In this paper, the learning rules of MFNNs are proposed and their convergence properties are studied. The results here show that the learning rules presented are convergent, which provides solid theoretical foundation for MFNNs’ application.