| 一种基于粗糙集理论的神经网络分类器的设计 |
| |
| 引用本文： | 李铁鹰,崔艳.一种基于粗糙集理论的神经网络分类器的设计[J].计算机工程与应用,2005,41(32):167-168,192. |
| |
| 作者姓名： | 李铁鹰 崔艳 |
| |
| 作者单位： | 北京理工大学,北京,100081;太原理工大学,太原,030024;太原理工大学,太原,030024 |
| |
| 摘 要： | 文章设计了一个基于粗糙集理论的神经网络分类器。该分类器利用粗糙集理论对原始数据进行特征选择,约简了冗余特性,减小了BP网络的输入维数,提高了网络的学习效率。在对一组数据实际分类的过程中,与单纯的神经网络分类器比较,在同等精度要求的情况下该分类器网络训练时间短,识别能力强。
|
| 关 键 词： | 粗糙集 BP神经网络 特性选择 分类 |
| 文章编号： | 1002-8331-（2005）32-0167-02 |
| 收稿时间： | 2005-01 |
| 修稿时间： | 2005-01 |
A Design of Neural Classifier Based on Rough Sets |
| |
| Li Tieying,Cui Yan.A Design of Neural Classifier Based on Rough Sets[J].Computer Engineering and Applications,2005,41(32):167-168,192. |
| |
| Authors: | Li Tieying Cui Yan |
| |
| Abstract: | A RS-based neural classifier is given in this paper.It selects the features from the original data by rough sets and reduces the redundant attributes.This can decrease the dimension of the input of BP neural network and improve the leaning efficiency of the network.In comparison to the common neural classifier,this one can make that training time is shortened and the ability of recognition is strengthened at the circumstance of equal precision required while a group of data is classified. |
| |
| Keywords: | rough sets BP Neural Network feature selection classification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
