| 一种混合混沌神经网络训练集的选择方法 |
| |
| 引用本文： | 祝先超,张新政.一种混合混沌神经网络训练集的选择方法[J].计算机与现代化,2009(1). |
| |
| 作者姓名： | 祝先超 张新政 |
| |
| 作者单位： | 广东工业大学自动化学院,广东,广州,510090 |
| |
| 基金项目： | 国家自然科学基金，广东省自然科学基金，广东省科技计划 |
| |
| 摘 要： | 阐述了一种利用欧氏距离和角距离选择混沌神经网络训练集的方法,该方法在欧氏距离的基础上,进一步利用角度最小原则,选择与参考中心角度最接近的部分,并通过模变换使其符合训练集的选择标准,实验证明其效果更佳.
|
| 关 键 词： | 混沌 神经网络 训练集 |
A Method to Choose Training Set of Mixed Chaos Neural Network |
| |
| ZHU Xian-chao,ZHANG Xin-zheng.A Method to Choose Training Set of Mixed Chaos Neural Network[J].Computer and Modernization,2009(1). |
| |
| Authors: | ZHU Xian-chao ZHANG Xin-zheng |
| |
| Affiliation: | School of Automation;Guangdong University of Technology;Guangzhou 510090;China |
| |
| Abstract: | A method to choose the training set of chaos neural network by using Euclidean distance and angular distance is expounded,the method used the principle of least angle furtherly,based on the Euclidean distance,choosed the part nearest to the angle of reference center,and it's accorded with the choice criterion of the training set through modular transformation,the function of the method is proved to be better by experimentation. |
| |
| Keywords: | chaos neural network training set |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
