Euclidean input mapping in a N-tuple approximation network

Author

Kolcz, Alek ; Allinson, Nigel M.

Author_Institution

Dept. of Electron., York Univ., UK

fYear

1994

fDate

2-5 Oct 1994

Firstpage

285

Lastpage

289

Abstract

A type of the N-tuple neural architecture can be shown to perform function approximation based on local interpolation, similar that performed by RBF networks. Since the size and speed of operation in this implementation are independent of the training set size, it is attractive for practical adaptive solutions. However, the kernel function used by the network is non-Euclidean, which can cause performance losses for high-dimensional input data. The authors investigate methods for realising more isotropic kernel basis functions by use of special data encoding techniques

Keywords

function approximation; interpolation; neural nets; Euclidean input mapping; N-tuple approximation network; function approximation; high-dimensional input data; isotropic kernel basis functions; kernel function; local interpolation; neural architecture; performance losses; special data encoding techniques; training set size; Electronic mail; Function approximation; Intelligent networks; Interpolation; Kernel; Laboratories; Performance loss; Radial basis function networks; Retina; Sampling methods;

fLanguage

English

Publisher

ieee

Conference_Titel

Digital Signal Processing Workshop, 1994., 1994 Sixth IEEE

Conference_Location

Yosemite National Park, CA

Print_ISBN

0-7803-1948-6

Type

conf

DOI

10.1109/DSP.1994.379821

Filename

379821