摘要
文中首先讨论了Ｂ样条基函数的特性，在此基础上采用构造性的方法从理论上证明了Ｂ样条神经网络能够以任意精度逼近任意定义在致密区间上的连续实函数．最后给出了构造性算法，使用此算法，能在满足误差要求的条件下，构造出几乎最小的Ｂ样条基函数．
The characteristics of B\|spline basis function is first discussed, based on which B\|spline neural network is proved to be a universal approximator. And then a constructive algorithm is presented. It is proved that this algorithm can be used to build a B\|spline neural network with minimum hidden units to approximate any continuous function defined on compact set to a prescribed accuracy.
出处
《计算机研究与发展》
EI
CSCD
北大核心
1999年第5期534-540,共7页
Journal of Computer Research and Development
基金
国家自然科学基金
