A Modified Density Based Outlier Mining Algorithm for Large Dataset

Outlier mining is to discover the objects with exceptional behavior in dataset. It is an important challenge from the knowledge discovery standpoint and attracts much attention recently. The density based outlier mining algorithm is an effective approach to detect anomalous points. However, such algorithms usually need amounts of computations. In this paper, we propose a modified density based detection algorithm which utilizes the data partitioning method. Furthermore, it presents some speedup strategies such as the introduction of module information to avoid large number of unnecessary computations while finding outliers. The algorithm is applied on both synthetic and real datasets and the experimental results show that it is efficient for outlier detection in large dataset.