| 噪声数据流的分类方法 |
| |
| 引用本文： | 陈丙杰,王晓晔,常飞.噪声数据流的分类方法[J].天津理工大学学报,2011,27(3):37-41. |
| |
| 作者姓名： | 陈丙杰 王晓晔 常飞 |
| |
| 作者单位： | 天津理工大学计算机与通信工程学院,天津,300384 |
| |
| 摘 要： | 数据流中噪声数据的处理是当前数据流分类挖掘中重要的研究分支,近些年来得到了广泛的关注.本文提出了一种称为FDBCA的数据流分类算法.它使用基于密度的带有噪声的空间聚类(DBSCAN)的改进算法Fast-DB-SCAN(FDBSCAN)处理噪声数据,并利用错误率方差(MSE)来检测概念漂移.同已有的数据流分类算法相比,实验结果表明了FDBCA算法可以提高噪声数据流的分类精度.
|
| 关 键 词： | 数据流 概念漂移 FDBSCAN 分类算法 |
Classification method of noise data stream |
| |
| CHEN Bing-jie,WANG Xiao-ye,CHANG Fei.Classification method of noise data stream[J].Journal of Tianjin University of Technology,2011,27(3):37-41. |
| |
| Authors: | CHEN Bing-jie WANG Xiao-ye CHANG Fei |
| |
| Institution: | (School of Computer and Communications Engineering,Tianjin University of Technology,Tianjin 300384,China) |
| |
| Abstract: | Processing noise data from data stream is one of the most important fields in data stream classification.In this paper,a new classification algorithm called FDBCA is proposed,which deals with noise data by Fast-DBSCAN that improved density based spatial clustering of applications with noise algorithm,and use mean square error to detect concept drifting.The experimental results show that FDBCA can improve classification accuracy on classifying noise data stream compared with existed data stream classification algorithm. |
| |
| Keywords: | data stream concept drifting FDBSCAN classification algorithm |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
