Dimensionality-Aware Outlier Detection: Theoretical and Experimental Analysis

arXiv:2401.05453v2 Announce Type: replace
Abstract: We present a nonparametric method for outlier detection that takes full account of local variations in intrinsic dimensionality within the dataset. Using the theory of Local Intrinsic Dimensionality (LID), our 'dimensionality-aware' outlier detection method, DAO, is derived as an estimator of an asymptotic local expected density ratio involving the query point and a close neighbor drawn at random. The dimensionality-aware behavior of DAO is due to its use of local estimation of LID values in …

abstract analysis arxiv cs.ai cs.lg dao dataset detection dimensionality estimator experimental intrinsic outlier theory type