An Effective Pruning based Outlier Detection Method to Quantify the Outliers

Outliers are the data objects that do not conform to the normal behaviour and usually deviates from the remaining data objects may be due to some outlying property which distinguishes them from the whole dataset. Usually, the detection of outliers is followed by the clustering of the dataset which sometimes ignores the prominency of outliers. In this study, we have tried to detect the outliers and pruned the clustering elements initially so that the outliers can be prominently highlighted. We have proposed an algorithm which effectively prunes the similar data objects from the large datasets and its experimental results compare the neighbouring points and show the better performance than the existing methods.