Clustering-Based Outlier Detection Method

Abstract

Outlier detection is important in many fields. The concept about outlier factor of object is extended to the case of cluster. Based on outlier factor of cluster, a clustering-based outlier detection method, named CBOD, is presented. The method consists of two stages, the first stage cluster dataset by one-pass clustering algorithm and second stage determine outlier cluster by outlier factor. The time complexity of CBOD is nearly linear with the size of dataset and the number of attributes, which results in good scalability and adapts to large dataset. The theoretic analysis and the experimental results show that the detection method is effective and practicable.