Enhancing effectiveness of density-based outlier mining scheme with density-similarity-neighbor-based outlier factor

摘要

This paper proposes a density-similarity-neighbor-based outlier mining algorithm for the data preprocess of data mining technique. First, the concept of k-density of an object is presented and the similar density series (SDS) of the object is established based on the changes of the k-density and the neighbors k-densities of the object. Second, the average series cost (ASC) of the object is obtained based on the weighted sum of the distance between the two adjacent objects in SDS of the object. Finally, the density-similarity-neighbor-based outlier factor (DSNOF) of the object is calculated by using both the ASC of the object and the ASC of k-distance neighbors of the object, and the degree of the object being an outlier is indicated by the DSNOF. The experiments are performed on synthetic and real datasets to evaluate the effectiveness and the performance of the proposed algorithm. The experiments results verify that the proposed algorithm has higher quality of outlier mining and do not increase the algorithm complexity.