Title :
A norm selection criterion for the generalized delta rule
Author :
Burrascano, Pietro
Author_Institution :
INFO-COM Dept., Rome Univ., Italy
fDate :
1/1/1991 12:00:00 AM
Abstract :
The derivation of a supervised training algorithm for a neural network implies the selection of a norm criterion which gives a suitable global measure of the particular distribution of errors. The author addresses this problem and proposes a correspondence between error distribution at the output of a layered feedforward neural network and Lp norms. The generalized delta rule is investigated in order to verify how its structure can be modified in order to perform a minimization in the generic Lp norm. The particular case of the Chebyshev norm is developed and tested
Keywords :
error statistics; learning systems; neural nets; Δ rule; Chebyshev norm; error distribution measure; generalized delta rule; generic Lp norm; layered feedforward neural network; minimization; norm selection criterion; supervised training algorithm; Backpropagation algorithms; Dispersion; Feedforward neural networks; Least squares methods; Multi-layer neural network; Multilayer perceptrons; Neural networks; Particle measurements; Stochastic processes; Testing;
Journal_Title :
Neural Networks, IEEE Transactions on
