Local Outlier Coefficient-Based Clustering Algorithm

This paper presents a local outlier coefficient-based clustering (LOCBC) algorithm. The algorithm introduces a new computation method about local outlier coefficient and the number of scanning the dataset is less than that of relative density based k-nearest neighbors (RDBKNN) clustering algorithm on the relative density computation. The experimental results show that LOCBC algorithm can effectively discover clusters of arbitrary shapes and outliers or noises. It can get good cluster quality and is more efficient than RDBKNN clustering algorithm