|
|||||
|
|
| Fourier三角基神经元网络的权值直接确定法 | |
| 引用本文： | 张雨浓,旷章辉,肖秀春,陈柏桃.Fourier三角基神经元网络的权值直接确定法[J].计算机工程与科学,2009,31(5). |
| 作者姓名： | 张雨浓 旷章辉 肖秀春 陈柏桃 |
| 作者单位： | 1. 中山大学信息科学与技术学院,广东,广州,510275 2. 中山大学信息科学与技术学院,广东,广州,510275;广东海洋大学信息学院,广东,湛江,524088 3. 广东海洋大学航海教育质管办,广东,湛江,524088 |
| 基金项目： | 国家自然科学基金，中山大学科研启动费、后备重点课题 |
| 摘 要： | 根据Fourier变换理论,本文构造出一类基于三角正交基的前向神经网络模型。该模型由输入层、隐层、输出层构成,其输入层和输出层采用线性激励函数,以一组三角正交基为其隐层神经元的激励函数。依据误差回传算法(即BP算法),推导了权值修正的迭代公式。针对BP迭代法收敛速度慢、逼近目标函数精度较低的缺点,进一步提出基于伪逆的权值直接确定法,该方法避免了权值反复迭代的冗长过程。仿真和预测结果表明,该方法比传统的BP迭代法具有更快的计算速度和更高的仿真与测试精度。 |
| 关 键 词： | 三角正交基函数 Fourier三角基神经元网络 权值修正 直接确定法 |
A Direct-Weight-Determination Method for Trigonometrically-Activated Fourier Neural Networks |
|
| ZHANG Yu-nong,KUANG Zhang-hui,XIAO Xiu-chun,CHEN Bo-tao.A Direct-Weight-Determination Method for Trigonometrically-Activated Fourier Neural Networks[J].Computer Engineering & Science,2009,31(5). | |
| Authors: | ZHANG Yu-nong KUANG Zhang-hui XIAO Xiu-chun CHEN Bo-tao |
| Affiliation: | 1.School of Information Science and Technology;Sun Yat-Sen University;Guangzhou 510275;2.School of Information;Guangdong Ocean University;Zhanjiang 524088;3.Maritime Education Quality System Office;China |
| Abstract: | Based on the Fourier transformation theory,a feed-forward neural network using trigonometric orthogonal activation-functions is constructed in this paper.The neural network adopts a three-layer structure,where the input and output layers employ linear activation functions,while the hidden-layer neurons are activated by a series of trigonometric orthogonal functions.In this paper,we first derive its weight-updating formula by adopting the standard BP training algorithm.More importantly,a pseudo-inverse metho... |
| Keywords: | trigonometric activation function trigonometrically-activated Fourier neural network weight-updating formula direct determination method |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |