摘要
研究多维Cardaliguet-Eurrard型神经网络算子的逼近问题.分别给出该神经网络算子逼近连续函数与可导函数的速度估计,建立了Jackson型不等式.
This paper deals with the approximation for multidimensional Cardaliguet- Eurrard type operators. The estimates of rates of approximation of these operators approximating the continuous functions and their derivatives are given, respectively. Also, two Jackson-type inequalities are established.
出处
《数学学报（中文版）》
SCIE
CSCD
北大核心
2008年第6期1161-1174,共14页
Acta Mathematica Sinica：Chinese Series
基金
国家自然科学基金资助项目(60473034，60575045)
浙江省教育厅科研重点基金资助项目(20060543)
关键词
神经网络
算子
逼近
neural networks
operator
approximation
