4096 search results - page 119 / 820 » Fibring Neural Networks |
137
click to vote
IJON
2008
15 years 4 months ago
2008
The term neural network evolution usually refers to network topology evolution leaving the network's parameters to be trained using conventional algorithms. In this paper we ...
128
click to vote
IPPS
2006
IEEE
15 years 10 months ago
2006
IEEE
This work presents a new approach to solve the location management problem by using the location areas approach. A combination of a genetic algorithm and the Hopfield neural netwo...
139
Voted
NN
2000
Springer
15 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, ) ...
120
click to vote
GECCO
2005
Springer
15 years 10 months ago
2005
Springer
Recurrent neural networks are theoretically capable of learning complex temporal sequences, but training them through gradient-descent is too slow and unstable for practical use i...
161
click to vote
APIN
2005
15 years 4 months ago
2005
Credit analysts generally assess the risk of credit applications based on their previous experience. They frequently employ quantitative methods to this end. Among the methods used...