Copula based approach for measurement validity verification in wireless sensor networks - Université de Bretagne Occidentale
Conference Papers Year : 2017

Copula based approach for measurement validity verification in wireless sensor networks

Sanaa Kawther Ghalem
  • Function : Author
  • PersonId : 1003927
Bouabdellah Kechar
  • Function : Author
  • PersonId : 885364
Ahcène Bounceur
Reinhardt Euler
Farid Lalem
  • Function : Author
  • PersonId : 10673
  • IdHAL : farid-lalem

Abstract

Outlier detection is the process of identifying the data objects that do not comply with the normal behavior of the defined data model. Used in automated data analysis, it ensures the desired data quality and reliability. This field has attracted increasing attention in the wireless sensor network domain, using methods from machine learning, data mining, and statistics literature. In this paper, we propose a novel outlier detection approach based on Copula theory. This powerful theory enables modeling the dependency between data measurements in a formal and statistical way.We have evaluated our proposed approach with a real world dataset. Our results show a detection rate of 85.90% and an error rate of 0.87% .
No file

Dates and versions

hal-01487535 , version 1 (12-03-2017)

Identifiers

  • HAL Id : hal-01487535 , version 1

Cite

Sanaa Kawther Ghalem, Bouabdellah Kechar, Ahcène Bounceur, Reinhardt Euler, Hammoudeh Mohammad, et al.. Copula based approach for measurement validity verification in wireless sensor networks. Second International Conference on Internet of Things, Data and Cloud Computing (ICC 2017), Mar 2017, Cambridge, United Kingdom. ⟨hal-01487535⟩
309 View
0 Download

Share

More