DocumentCode :
984161
Title :
Numerical integration based on a neural network algorithm
Author :
Zhe-zhao, Zeng ; Yao-nan, Wang ; Hui, Wen
Author_Institution :
Hunan Univ., Changsha, China
Volume :
8
Issue :
4
fYear :
2006
Firstpage :
42
Lastpage :
48
Abstract :
The authors present a neural network approach to numerical integration and propose four methods for training neural networks. Their results show that their own numerical integration method has high accuracy and a fast convergence rate, making its application values significant for engineering practice.
Keywords :
integration; neural nets; convergence rate; engineering practice; neural network algorithm; numerical integration; Convergence of numerical methods; Differential equations; Digital control; Gaussian processes; Interpolation; Large-scale systems; Neural networks; Polynomials; Proportional control; Three-term control; convergence rates; neural networks; numerical integration;
fLanguage :
English
Journal_Title :
Computing in Science & Engineering
Publisher :
ieee
ISSN :
1521-9615
Type :
jour
DOI :
10.1109/MCSE.2006.73
Filename :
1644707
Link To Document :
بازگشت