摘要
在［３］的基础上，本文给出了基于属性整合的模式生成与转化机制的数学表达，并对跨感觉模式识别中的特征辨识和模式匹配、补充和修正问题。
A mathematical model of sensation neuron detection, which is based on attributive abstraction and integration, is presented in this paper. It is shown that both spatial frequency analysis theory and feature abstraction theory can be uniformed in this model, and it is not only an attributive semantic extension of artificial neural networks, but also an integration of symbol mechanism and network mechanism.
出处
《计算机研究与发展》
EI
CSCD
北大核心
1997年第7期481-486,共6页
Journal of Computer Research and Development
基金
国家自然科学基金
国家"863"高技术计划基金
国家攀登计划基金
广西自然科学基金
关键词
属性抽取与整合
感觉检测
神经网络
attributive abstraction and integration, sensation detection, neural networks
