3 search results - page 1 / 1 » Best approximation by Heaviside perceptron networks |
NN
2000
Springer
13 years 4 months ago
2000
Springer
In Lp-spaces with p [1, ) there exists a best approximation mapping to the set of functions computable by Heaviside perceptron networks with n hidden units; however for p (1, ) ...
CVPR
2012
IEEE
11 years 7 months ago
2012
IEEE
Image denoising can be described as the problem of mapping from a noisy image to a noise-free image. The best currently available denoising methods approximate this mapping with c...
IJON
2007
13 years 4 months ago
2007
Piecewise linear networks (PLNs) are attractive because they can be trained quickly and provide good performance in many nonlinear approximation problems. Most existing design alg...