Article ID Journal Published Year Pages File Type
405556 Neural Networks 2011 9 Pages PDF
Abstract

Multilayer perceptron networks whose outputs consist of affine combinations of hidden units using the tanhtanh activation function are universal function approximators and are used for regression, typically by reducing the MSE with backpropagation. We present a neural network weight learning algorithm that directly positions the hidden units within input space by numerically analyzing the curvature of the output surface. Our results show that under some sampling requirements, this method can reliably recover the parameters of a neural network used to generate a data set.

Related Topics
Physical Sciences and Engineering Computer Science Artificial Intelligence
Authors
, , , ,