Ellipsoidal neighbourhood outlier factor for distributed anomaly detection in resource constrained networks
作者：
Highlights：
• Detecting anomalies in data is challenging on resource constrained networks.
• A hyperEllipsoidal Neighborhood Outlier Factor (ENOF) is proposed.
• A distributed algorithm using hypersellipsoidal clusters and ENOF scheme is proposed.
• Capable of identifying local and global anomalies at individual node levels.
• Achieves superior detection capabilities with minimal communication overhead.
摘要
Highlights•Detecting anomalies in data is challenging on resource constrained networks.•A hyperEllipsoidal Neighborhood Outlier Factor (ENOF) is proposed.•A distributed algorithm using hypersellipsoidal clusters and ENOF scheme is proposed.•Capable of identifying local and global anomalies at individual node levels.•Achieves superior detection capabilities with minimal communication overhead.
论文关键词：Anomaly detection,Outlier factor,Hyperellipsoidal model,Distributed detection,Sensor networks
论文评审过程：Received 24 July 2013, Revised 6 March 2014, Accepted 1 April 2014, Available online 12 April 2014.
论文官网地址：https://doi.org/10.1016/j.patcog.2014.04.006
