Dimensionality-Aware Outlier Detection: Theoretical and Experimental Analysis

Alastair Anderberg, James Bailey, Ricardo J. G. B. Campello, Michael E. Houle, Henrique O. Marques +2 more
1/10/2024
cs.LGcs.AI

No official code found. Best implementation: homarques/DAO (4 stars, last commit 2024-10-15).

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 a theoretically-justified way. Through comprehensive experimentation on more than 800 synthetic and real datasets, we show that DAO significantly outperforms three popular and important benchmark outlier detection methods: Local Outlier Factor (LOF), Simplified LOF, and kNN.

View on arXivView PDF

Code Implementations(1)

homarques/DAOCommunity100%

Dimensionality-Aware Outlier Detection

41R, PythonJun 13, 20232 years ago

Similar Papers

DCFO Additional Material

83% match2025

Comparative Study of Neighbor-based Methods for Local Outlier Detection

82% match2024

DCFO: Density-Based Counterfactuals for Outliers - Additional Material

82% match2025

Outlier Detection Using Vector Cosine Similarity by Adding a Dimension

82% match2025

Unsupervised Parameter-free Outlier Detection using HDBSCAN* Outlier Profiles

81% match1 repo22024

An Efficient Outlier Detection Algorithm for Data Streaming

81% match2025

Cite this paper

@article{anderberg2024dimensionalityaware,
  title  = {Dimensionality-Aware Outlier Detection: Theoretical and Experimental Analysis},
  author = {Alastair Anderberg and James Bailey and Ricardo J. G. B. Campello and Michael E. Houle and Henrique O. Marques and Miloš Radovanović and Arthur Zimek},
  year   = {2024},
  eprint = {2401.05453},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2401.05453v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2401.05453.svg)](https://codeofpaper.com/paper/2401.05453)

Discussion