Parameterless Outlier Detection in Data Streams

Alice Marascu 1 Florent Masseglia 1
1 AxIS - Usage-centered design, analysis and improvement of information systems
CRISAM - Inria Sophia Antipolis - Méditerranée , Inria Paris-Rocquencourt
Abstract : Outlyingness is a subjective concept relying on the isolation level of a (set of) record(s). Clustering-based outlier detection is a field that aims to cluster data and to detect outliers depending on their characteristics (small, tight and/or dense clusters might be considered as outliers). Existing methods require a parameter standing for the "level of outlyingness", such as the maximum size or a percentage of small clusters, in order to build the set of outliers. Unfortunately, manually setting this parameter in a streaming environment should not be possible, given the fast time response usually needed. In this paper we propose WOD, a method that separates outliers from clusters thanks to a natural and effective principle. The main advantages of WOD are its ability to automatically adjust to any clustering result and to be parameterless.
Document type :
Conference papers
Complete list of metadatas

https://hal.inria.fr/inria-00461827
Contributor : Alice-Maria Marascu <>
Submitted on : Friday, March 5, 2010 - 5:51:21 PM
Last modification on : Saturday, February 23, 2019 - 7:06:02 PM

Identifiers

  • HAL Id : inria-00461827, version 1

Collections

Citation

Alice Marascu, Florent Masseglia. Parameterless Outlier Detection in Data Streams. ACM symposium on Applied Computing, Mar 2009, Honolulu, United States. pp.1491-1495. ⟨inria-00461827⟩

Share

Metrics

Record views

180