Similarity-Based Unsupervised Evaluation of Outlier Detection

Henrique O. Marques*, Arthur Zimek, Ricardo J.G.B. Campello, Jörg Sander

*Kontaktforfatter

Publikation: Kapitel i bog/rapport/konference-proceedingKonferencebidrag i proceedingsForskningpeer review

Abstract

The evaluation of unsupervised algorithm results is one of the most challenging tasks in data mining research. Where labeled data are not available, one has to use in practice the so-called internal evaluation, which is based solely on the data and the assessed solutions themselves. In unsupervised cluster analysis, indices for internal evaluation of clustering solutions have been studied for decades, with a multitude of indices available, based on different criteria. In unsupervised outlier detection, however, this problem has only recently received some attention, and still very few indices are available. In this paper, we provide a new internal index based on criteria different from the ones available in the literature. The index is based on a (generic) similarity measure to efficiently evaluate candidate outlier detection solutions in a completely unsupervised way. We evaluate and compare this index against existing indices in terms of quality and run time performance using collections of both real and synthetic datasets.

OriginalsprogEngelsk
TitelSimilarity Search and Applications - 15th International Conference, SISAP 2022, Proceedings
RedaktørerTomáš Skopal, Jakub Lokoč, Fabrizio Falchi, Maria Luisa Sapino, Ilaria Bartolini, Marco Patella
ForlagSpringer Science+Business Media
Publikationsdato2022
Sider234-248
ISBN (Trykt)9783031178481
DOI
StatusUdgivet - 2022
Begivenhed15th International Conference on Similarity Search and Applications, SISAP 2022 - Bologna, Italien
Varighed: 5. okt. 20227. okt. 2022

Konference

Konference15th International Conference on Similarity Search and Applications, SISAP 2022
Land/OmrådeItalien
ByBologna
Periode05/10/202207/10/2022
NavnLecture Notes in Computer Science
Vol/bind13590 LNCS
ISSN0302-9743

Bibliografisk note

Funding Information:
Acknowledgement. This work has partly been funded by NSERC Canada, and the Independent Research Fund Denmark in the project “Reliable Outlier Detection”.

Publisher Copyright:
© 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Fingeraftryk

Dyk ned i forskningsemnerne om 'Similarity-Based Unsupervised Evaluation of Outlier Detection'. Sammen danner de et unikt fingeraftryk.

Citationsformater