Kůrková V, Savický P, Hlavácková K
Institute of Computer Science, Academy of Sciences of the Czech Republic, Pod vodárensku vezi; 2, P.O. Box 5 182 07, Prague, Czech Republic
Neural Netw. 1998 Jun;11(4):651-659. doi: 10.1016/s0893-6080(98)00039-2.
We give upper bounds on rates of approximation of real-valued functions of d Boolean variables by one-hidden-layer perceptron networks. Our bounds are of the form c/n where c depends on certain norms of the function being approximated and n is the number of hidden units. We describe sets of functions where these norms grow either polynomially or exponentially with d.
我们给出了由单隐藏层感知器网络对d个布尔变量的实值函数进行逼近的速率的上界。我们的界具有c/n的形式,其中c取决于被逼近函数的某些范数,n是隐藏单元的数量。我们描述了这些范数随d呈多项式增长或指数增长的函数集。