Let 1 ≤ c ≤ d and let g be a slowly increasing function defined on Rc. Suppose that the support of the Fourier transform Fcg of g includes a converging sequence of distinct points yκ which sufficiently rapidly come close to a line as κ→ ∞. Then, any mapping of any number, say n, of any points x1, xn in Rd onto R can be implemented by a linear sum of the form ∑ n j=1 ajg(Wxi + zj ). Here, W is a d × c matrix having orthonormal row vectors, implying that g is used without scaling, and that the sigmoid function defined on R and the radial basis function defined on Rd are treated on a common basis. © Springer-VerlagBerlin Heidelberg 2002.
CITATION STYLE
Ito, Y. (2002). A weak condition on linear independence of unscaled shifts of a function and finite mappings by neural networks. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2415 LNCS, pp. 337–343). Springer Verlag. https://doi.org/10.1007/3-540-46084-5_55
Mendeley helps you to discover research relevant for your work.