An outlier detection method based on cluster pruning

Abstract

Outlier detection has a wide range of applications. In this paper we present a new method for detecting outliers, focused on reducing the number of computations. Our method operates on two phases and uses one pruning strategy. Objective is to remove the points which are considered to be inliers. In the first phase a clustering algorithm is applied to partition the data into clusters and make an estimate to prune the clusters, in the second phase we apply a outlier score function to dictate the outliers. The experimental results using real datasets demonstrate the superiority of our method over existing outlier detection method.