Pruning based method for outlier detection

In this paper we propose a method to capture outliers. We apply a clustering algorithm to divide the dataset into independent clusters. The clusters which are dense in nature doesnot contain outliers. And the clusters which are sparse are probable candidate clusters for outliers. Pruning the dense clusters makes the dataset small and sparse. For the unpruned points we calculated a distance based outlier score. The computations needed for calculating the outlier score reduces considerably due to the pruning of many points. Based on the outlier score we declare the top-n points with the highest score as outliers. The experimental results using real data set demonstrate that even though the number of computations are less, the proposed method performs better than the existing method.