scientific article; zbMATH DE number 1193393
From MaRDI portal
Publication:3841869
Large-dimensional wavelets are much less often studied than one- or two-dimensional wavelets, due to practical difficulties for the implementation of large-dimensional wavelet bases or frames. A method for constructing large-dimensional wavelet frames is presented in this paper. A radial construction of such frames is proposed. The use of large-dimensional wavelet frames in one hidden layer neural networks is also considered, as a practically reasonable application. Such networks are particular useful for nonlinear function approximations. For the entire collection see [Zbl 0889.00018].
Recommendations
Cited in
(2)
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3841869)