Enhancing Effectiveness of Density-Based Outlier Mining

Outlier mining is an important work of data mining and a density-similarity-neighbor based outlier factor (DSNOF) algorithm is proposed to indicate the degree of outlier-ness of an object. The proposed algorithm calculates the densities of an object and its neighbors and constructs the similar density series (SDS) in the neighborhood of the object. Based on the SDS, the proposed algorithm computes the average series cost (ASC) of the object and the DSNOF of the object can be obtained according to the ASC of the object and those of the neighbors of the object. The experiments are performed on the synthetic and the real datasets. The experiments results verify that the proposed algorithm not only can detect outlier more effectively and but also do not increase the time and the space complexities.