计算机与现代化 ›› 2010, Vol. 1 ›› Issue (01): 32-36.doi:

• 算法分析与设计 • 上一篇    下一篇

数据流分类技术分析

谢琰嵩,朱晓莲   

  1. 中国地质大学(武汉)计算机学院,湖北 武汉 430074
  • 收稿日期:2008-12-19 修回日期:1900-01-01 出版日期:2010-01-15 发布日期:2010-01-15

Analysis of Classification Technology of Data Stream

XIE Yan-song,ZHU Xiao-lian   

  1. School of Computer Science, China University of Geosciences(Wuhan), Wuhan 430074, China
  • Received:2008-12-19 Revised:1900-01-01 Online:2010-01-15 Published:2010-01-15

摘要: 随着科学技术发展的不断进步,静态数据上的分类技术已不能满足现实情况的需要了。大量的数据都是以数据流的形式出现,本文对这种新数据形式下的分类算法进行了一个综述。所描述的主要算法有:集合分类器、基于需求的数据流分类算法、最近邻数据流分类算法、快速决策树算法。通过学习研究和实验对比结果发现,这些数据流上的分类算法性能完全超过了传统的分类算法。

关键词: 数据流, 分类器, 概念漂移

Abstract: With the continuous advancement of scientific and technological development, static data classification technology can not meet the real needs of the situation. Large amounts of data are based on data stream, this paper summarizes the new form of data classification algorithms. The main algorithms which being described are ensemble classification, ANNCAD and VFDT. Through the study of research and experimental comparison of the results, people can find these data streams on classification algorithms performance far exceeds the traditional classification algorithms.

Key words: data stream, classification, concept drift

中图分类号: