Sciweavers

4 search results - page 1 / 1
» Solving the XOR and parity N problems using a single univers...
Sort
View
VLSISP
2002
114views more  VLSISP 2002»
13 years 4 months ago
Image processing using cellular neural networks based on multi-valued and universal binary neurons
Multi-valued and universal binary neurons (MVN and UBN) are the neural processing elements with the complex-valued weights and high functionality. It is possible to implement an a...
Igor N. Aizenberg, Constantine Butakoff
ISNN
2005
Springer
13 years 10 months ago
Enhanced Fuzzy Single Layer Perceptron
Abstract. In this paper, a method of improving the learning time and convergence rate is proposed to exploit the advantages of artificial neural networks and fuzzy theory to neuron...
Kwang-Baek Kim, Sungshin Kim, Young Hoon Joo, Am S...
TNN
2010
234views Management» more  TNN 2010»
12 years 11 months ago
Novel maximum-margin training algorithms for supervised neural networks
This paper proposes three novel training methods, two of them based on the back-propagation approach and a third one based on information theory for Multilayer Perceptron (MLP) bin...
Oswaldo Ludwig, Urbano Nunes