Tuple Compress Based Outlier Detection on Uncertain Data of Mutually Exclusive Relation

Outlier detection techniques have widely been applied in medicine, finance, information security and so on. These techniques have been well studied on deterministic data. But, in some important application domains such as sensor networks, moving object tracking and data cleaning, uncertainty is inherent in data due to various factors. Furthermore, those uncertain data may have mutually exclusive relation. How to detect outliers on uncertain data of mutually exclusive relation is a new challenge. In this paper, a new definition of outlier on uncertain data is defined. A tuple compress based outlier detection method is proposed. Experimental results show that the proposed approach can efficiently detect outliers in data set.