Feed-forward network training using optimal input gains

Authors:
Sanjeev S. Malalur;Michael Manry
Affiliations:
Electrical Engineering Department, University of Texas at Arlington, TX;Electrical Engineering Department, University of Texas at Arlington, TX
Venue:
IJCNN'09 Proceedings of the 2009 international joint conference on Neural Networks
Year:
2009

Citing 5
Cited 0

Learning internal representations by error propagation

Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
Characterization of a class of sigmoid functions with applications to neural networks

Neural Networks
Data Mining with Computational Intelligence (Advanced Information and Knowledge Processing)

Data Mining with Computational Intelligence (Advanced Information and Knowledge Processing)
Neural-network classifiers for recognizing totally unconstrained handwritten numerals

IEEE Transactions on Neural Networks
Improving model accuracy using optimal linear combinations of trained neural networks

IEEE Transactions on Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, an effective batch training algorithm is developed for feed-forward networks such as the multilayer perceptron. First, the effects of input transforms are reviewed and explained, using the concept of equivalent networks. Next, a non-singular diagonal transform matrix for the inputs is proposed. Use of this transform is equivalent to altering the input gains in the network. Newton's method is used to solve for the input gains and an optimal learning factor. In several examples, it is shown that the final algorithm is a reasonable compromise between first order training methods and Levenburg-Marquardt.