›› 2015, Vol. 28 ›› Issue (1): 155-.

• 论文 • 上一篇    下一篇

数据流聚类算法在数据处理中的应用

江楠,徐秦   

  1. (中国人民解放军92124部队,辽宁 大连 116023)
  • 出版日期:2015-01-15 发布日期:2015-01-22
  • 作者简介:江楠(1982—),女,硕士,助理工程师。研究方向:通信工程。E-mail:66757588@qq.com。徐秦(1972—),女,高级工程师。研究方向:通信工程。

Feature-based Data Stream Clustering Algorithm Projection and Fitting in Data Processing

JIANG Nan,XU Qin   

  1. (92124 Troops of the Chinese People's Liberation Army,Dalian 116023,China)
  • Online:2015-01-15 Published:2015-01-22

摘要:

基于监测数据进行分析,采用特殊投影和拟合的数据流聚类算法,该算法结合拟用算法和滑动窗口技术进行聚类预处理,掌握数据流的变化趋势。此外,通过桥梁健康监测仿真实验,得出拟合数据具有更高的精确性,分段拟合的数据R2接近1,m为0.890 4<0.5,解决了高维数据流的降维问题,缩短了数据处理时间,且提高效率39.1以上,占用空间仅为6.7%。

关键词: 特殊投影, 拟合数据, 聚类算法, 滑动窗口

Abstract:

The trend of data stream is analyzed based on monitoring data using a special projection clustering and fitting algorithm,which combines the pre-processing algorithms and sliding window technique.The bridge health monitoring simulation experiments are performed with fitting data of higher accuracy (segment R2 at distance close to 1,m=0.890 4<0.5),thus solving the dimensionality reduction of high dimensional data stream,and reducing the data processing time with an increase in efficiency by 39.1% and a space occupation of only 6.7%.

Key words: special projection;fit the data;clustering algorithm;sliding window

中图分类号: 

  • TP301.6