An Efficient Learning Algorithm for Feedforward Neural Networks

杜正春; 刘玉田; 夏道止

您当前的位置：

首页 >

文章列表页 >

An Efficient Learning Algorithm for Feedforward Neural Networks

更新时间：2025-12-08

- An Efficient Learning Algorithm for Feedforward Neural Networks
- Acta Electronica Sinica Issue 8, (1995)
- 作者机构：
  
  西安交通大学
- 作者简介：
- 基金信息：
- DOI：
  CLC： TP301.6
- Published：1995
- 稿件说明：
移动端阅览
[1]杜正春,刘玉田,夏道止.前馈神经网络的一种有效学习算法[J].电子学报,1995(08):57-61.

杜正春, 刘玉田, 夏道止. An Efficient Learning Algorithm for Feedforward Neural Networks[J]. Acta Electronica Sinica, 1995, (8).
[1]杜正春,刘玉田,夏道止.前馈神经网络的一种有效学习算法[J].电子学报,1995(08):57-61. DOI：

杜正春, 刘玉田, 夏道止. An Efficient Learning Algorithm for Feedforward Neural Networks[J]. Acta Electronica Sinica, 1995, (8). DOI：

摘要

本文提出了基于混合ＧＮ－ＢＦＧＳ法进行前馈神经网络学习的新算法。该算法结合ＧＮ法与ＢＦＧＳ法的特点，既利用了问题本身的特殊结构，又能取得超线性甚至二次渐近收敛率。与ＢＰ算法相比，这种算法可取得更快和更可靠的学习特性，在学习过程中利用该方法能够区分非零残量和零残量问题的特点，提出了自动调整隐单元数的方法，从而可以保证网络的学习与归纳能力。示例系统的结果表明了所提方法的有效性。

Abstract

A new learning algorithm for feedforward neural networks based on the hybrid GNBFGS method is presented. The algorithm combines the better features of both Gauss-Newton and BFGS methods

which make use of the special structure of the problem

and the order of convergence is superlinear even quadratic.It gives faster and more reliable learning property compared with the BP algorithm.By making use of the feature for distinguishing the non-zero-residual and zero-residual problem of hybrid GN-BFGS method

a technique to adjust the number of hidden units in learning process is proposed to ensure learning and generalization capabilities of networks.The results of sample systems show the effectiveness of the algorithm.

关键词

Keywords

references

Views

128

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

High Attribute Dimensional Sparse Clustering Recurrent Logical Neural Networks Model and Learning Algorithm

On the Perceptron Convergence Theorem of the Monolithic Fuzzy Neural Networks

Analysis and Theory of High-Dimension Space Geometry for Artificial Neural Networks

A New Layer-Wise Linearized Algorithm for Feedforward Neural Network

Study on the Recognitions of Isolated Chinese Syllables by Neural Nets

Related Author

JIAO Li-cheng

ZHOU Yong-quan

HUANG De-shuang

HE Xin-gui

LIANG Jiu-zhen

WANG Bai-nan

WANG Shou-jue

TIAN Chuan-jun

Related Institution

College of Computer and Information ScienceGuangxi University for NationalitiesNanningGuangxi 530006China

Institute of Intelligence Information ProcessingXidian UniversityXi'anShaanxi 710071China

College of Computer and Information Science,Guangxi University for Nationalities

Institute of Intelligence Information Processing,Xidian University

Beijing Institute of System EngineeringBeijing 100101China

⁰