An Improved Overlapping Clustering Algorithm to Detect Outlier

Abstract

MCOKE algorithm in identifying data objects to multi cluster is known for its simplicity and effectiveness. Its drawback is the use of maxdist as a global threshold in assigning objects to one or more cluster while it is sensitive to outliers. Having outliers in the datasets can significantly affect the effectiveness of maxdist as regards to overlapping clustering. In this paper, the outlier detection is incorporated in MCOKE algorithm so that it can detect and remove outliers that can participate in the calculation of assigning objects to one or more clusters. The improved MCOKE algorithm provides better identification of overlapping clustering results. The performance was evaluated via F1 score performance criterion. Evaluation results revealed that the outlier detection demonstrated higher accuracy rate in identifying abnormal data (outliers) when applied to real datasets.