APPROXIMATION CAPABILITY OF LAYERED NEURAL NETWORKS WITH SIGMOID UNITS ON 2 LAYERS

Using only an elementary constructive method, we prove the universal approximation capability of three-layered feedforward neural networks that have sigmoid units on two layers. We regard the Heaviside function as a special case of sigmoid function and measure accuracy of approximation in either the supremum norm or in the L(p)-norm. Given a continuous function defined on a unit hypercube and the required accuracy of approximation, we can estimate the numbers of necessary units on the respective sigmoid unit layers. In the case where the sigmoid function is the Heaviside function, our result improves the estimation of Kurkova (1992). If the accuracy of approximation is measured in the L(p)-norm, our estimation also improves that of Kurkova (1992), even when the sigmoid function is not the Heaviside function.