| 二次型最小化所展现的BP与Hopfield类型神经网络的学习同质性 |
| |
| 引用本文： | 张雨浓,麦剑章,肖秀春,李展,易称福. 二次型最小化所展现的BP与Hopfield类型神经网络的学习同质性[J]. 自动化技术与应用, 2008, 27(9): 6-10 |
| |
| 作者姓名： | 张雨浓 麦剑章 肖秀春 李展 易称福 |
| |
| 作者单位： | 中山大学电子与通信工程系,广东,广州,510275;中山大学软件学院,广东,广州,510275 |
| |
| 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： | 本论文揭示,作为两种并行的神经计算模型,BP和Hopfield类型神经网络都可以有效地对二次型V（x）=x^TPx/2＋q^Tx实现最小化求解。而且,尽管BP和Hopfield类型神经网络在网络设计思想和网络结构上呈现出很大的差异,但是它们在二次型函数最小化问题上都表现出了相同的学习能力,这说明两者具有本质的联系.
|
| 关 键 词： | 二次型函数最小化 BP神经网络 Hopfield类型神经网络 学习同质性 |
The Common Nature of Learning of the BP and Hopfield-type Neural Networks in the Minimization of Quadratic Functions |
| |
| ZHANG Yun-ong,MAI Jian-zhang,XIAO Xiu-chun,LI Zhan,YI Chen-fu. The Common Nature of Learning of the BP and Hopfield-type Neural Networks in the Minimization of Quadratic Functions[J]. Techniques of Automation and Applications, 2008, 27(9): 6-10 |
| |
| Authors: | ZHANG Yun-ong MAI Jian-zhang XIAO Xiu-chun LI Zhan YI Chen-fu |
| |
| Affiliation: | ZHANG Yu-nong, MAI Jian-zhang, XIAO Xiu-chun, LI Zhan, YI Chen-fu (1. Department of Electronics and Communication Engineering, Sun Yat-Sen University, Guangzhou 510275 China; 2. School of Software, Sun Yat-Sen University, Guangzhou 510275 China) |
| |
| Abstract: | As parallel-computational models, both BP （Back Propagation） and Hopfield-type neural networks can be used in the minimization of quadratic functions. BP neural network is substantially different from Hopfield-type neural network in terms of network architecture and learning pattern. However, both neural networks possess a common nature of learning during online minimization of quadratic functions. Simulation results are also given. |
| |
| Keywords: | minimization of quadratic functions BP neural network Hopfield-type neural network common nature of learning |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
