摘要
提出了一种模糊神经元网络的学习算法即利用多层模糊IF/THEN规则表达专家知识的神经网络学习方法。在以此构造的基于多源信息融合的分类系统中,采用了多层模糊IF/THEN规则进行分类。为了处理模糊语言值,提出了一种能够控制模糊输入矢量的神经网络体系结构,该方法能够对非线性实间隔矢量和模糊矢量进行分类。工程实验表明,此学习算法是切实可行的。
A learning algorithm of fuzzy neural network has been presented in this paper ,namely, neural network which learns through fuzzy IF/THEN rules. IF/THEN rules and Digital Data are used in the classification system based on the multi-source information fusion. A fuzzy neural architecture is given to process fuzzy input vectors and output fuzzy value. A method that can clarify given non-linear real intervals or fuzzy vectors is given. From our experiments, we can know its feasibility.
出处
《计算机工程》
CAS
CSCD
北大核心
2002年第1期31-32,49,共3页
Computer Engineering
基金
国家自然科学基金项目(69873007)
