|Table of Contents|

Decentralized quickest change detection based on cluster analysis

《南京理工大学学报》(自然科学版)[ISSN:1005-9830/CN:32-1397/N]

Issue:
2014年02期
Page:
276-280
Research Field:
Publishing date:

Info

Title:
Decentralized quickest change detection based on cluster analysis
Author(s):
Wang Xin
Key Laboratory of Advanced Process Control for Light Industry,Ministry of Education, Jiangnan University,Wuxi 214122,China
Keywords:
decentralized detection quickest change detection cluster analysis principal component analysis
PACS:
TN911.23
DOI:
-
Abstract:
The decentralized quickest change detection is a common decision problem in the sensor network application.Being different from the classic quickest change detection based on the cumulative sum statistics,an idea of the decentralized quickest change detection based on the cluster analysis is proposed,and an implementation empirically feasible is designed.The implementation comes from the equivalence between K-means clustering and principal component analysis.The implementation does not require pre-change and post-change probability density function of the sensor observation values and the sensors'local computing in advance,and numerical simulations show that it has a smaller detection delay and a smaller probability of false alarm.

References:

[1] Teneketzis D.The decentralized quickest detection problem[A].Proceedings of the 21st IEEE Conference on Decision and Control[C].Orlando,US:IEEE Press,1982:673-679.
[2]Veeravalli V V.Decentralized quickest change detection[J].IEEE Transactions on Information Theory,2001,47(4):1657-1665.
[3]Mei Y.Efficient scalable schemes for monitoring a large number of data streams[J].Biometrika,2010,97(2):419-433.
[4]Mei Yajun.Quickest detection in censoring sensor networks[A].Proceedings(ISIT)2011 IEEE International Symposium on Information Theory[C].St Petersburg,Russia:IEEE Press,2011:2148-2152.
[5]王枚,王国宏.基于主成分分析的目标确认方法及其在车标定位识别中的应用[J].南京理工大学学报,2009,33(1):42-46.
Wang Mei,Wang Guohong.Object confirmation method based on principle component analysis and its application in vehicle-logo location and recognition[J].Journal of Nanjing University of Science and Technology,2009,33(1):42-46.
[6]Ding C,He X.K-means clustering via principal component analysis[A].Proceedings of the 21st International Conference on Machine Learning[C].Banff,Canada:ACM Bress,2004:225-232.
[7]Kulis B,Jordan M I.Revisiting K-means:New algorithms via Bayesian nonparametrics[A].Proceedings of the 29th International Conference on Machine Learning[C].Edinburgh,UK:Omnipress,2012:513-520.
[8]Ilin A,Raiko T.Practical approaches to principal component analysis in the presence of missing values[J].Journal of Machine Learning Research,2010,11:1957-2000.

Memo

Memo:
-
Last Update: 2014-04-30