摘要
本文讨论Ｅｄ中有限超平面集对Ｅｄ的划分问题，证明了划分中各种维数连通部分个数的上确界，分析超平面集对Ｅｄ的划分问题在模式识别、人工神经网络。
The dissection of H by a finite set H of hyperplanes in Ed is discussed.Tight upper bounds of the numbers of connected pieces of various dimensions are proved.Based on the analysis of the dissection of H,its applications in pattern recognition,artificial neural networks,computer vision,and computational learning theory are illustrated.
出处
《计算机学报》
EI
CSCD
北大核心
1994年第9期697-702,共6页
Chinese Journal of Computers
基金
863国家高技术资助
国家自然科学基金
关键词
模式识别
神经网络
划分
Pattern recognition,artificial neural network,aspect graph,computational learning theory.
