Realization of RBF Neural Network with Local and Semi-Local Transfer Function Approximation and Classification-Reference-Cited by-同舟云学术

Realization of RBF Neural Network with Local and Semi-Local Transfer Function Approximation and Classification

Published:2014-11 Issue: Volume:687-691 Page:1628-1632
ISSN:1662-7482
Container-title:Applied Mechanics and Materials
language:
Short-container-title:AMM

Author:

Luo Mao¹,Song Shao Yun¹

Affiliation:

1. Yuxi Normal University

Abstract

Incremental Neural Network (IncNet) structure is controlled by the growth and pruning, and the complexity of the match and training data. Dual radial transfer function is more flexible than other commonly transfer function used in artificial neural network. Recent improvements in the multi-dimensional space (having the N-1 parameters) to increase the rotation of the transfer function of the constant value. Based on the results of the benchmark approach and psychological classification analysis clearly shows than any other classification network model has a stronger generalization.

Publisher

Trans Tech Publications, Ltd.

Link

https://www.scientific.net/AMM.687-691.1628.pdf

Reference4 articles.

1. R. Adamczak, W. Duch, and N. Jankowski. New developments in the Feature Space Mapping model. In Third Conference on Neural Networks and Their Applications, pages 65–70, Kule, Poland, Oct. (1997).

2. W. Duch and N. Jankowski. Survey of neural transfer functions. Neural Computing Surveys, 7, 1999. (submitted).

3. F. Girosi. An equivalence between sparse approximation and support vector machines. Neural Computation, 10(6), Aug. (1998).

4. N. Jankowski. Ontogenic neural networks and their applications to classification of medical data. PhD thesis, Department of Computer Methods, Nicholas Copernicus University, Torun, Poland, 1999. (in preparation).