Title :
A simplified statistical analysis of feedforward neural classifier
Author :
Wang, Chuan ; Principe, Jose C.
Author_Institution :
Dept. of Electr. Eng., Florida Univ., Gainesville, FL, USA
fDate :
27 Jun-2 Jul 1994
Abstract :
We propose an analytical procedure to determine the parameters of a neural classifier. A cost function based on the input statistics is defined. We also suggest an approximation model to describe the output distribution of each nonlinear unit with Gaussian inputs. We show that the resulting parametric classifier can perform as well as a conventional neural classifier trained with the backpropagation algorithm to minimize the mean square error
Keywords :
feedforward neural nets; pattern classification; statistical analysis; Gaussian inputs; approximation model; backpropagation algorithm; cost function; feedforward neural classifier; mean square error minimization; parametric classifier; simplified statistical analysis; Backpropagation algorithms; Cost function; Density functional theory; Gaussian distribution; Neural engineering; Neural networks; Neurons; Parametric statistics; Statistical analysis; Statistical distributions;
Conference_Titel :
Neural Networks, 1994. IEEE World Congress on Computational Intelligence., 1994 IEEE International Conference on
Conference_Location :
Orlando, FL
Print_ISBN :
0-7803-1901-X
DOI :
10.1109/ICNN.1994.374282
