Anomaly Detection and Localisation using Mixed Graphical Models

Abstract : We propose a method that performs anomaly detection and localisation within heterogeneous data using a pairwise undirected mixed graphical model. The data are a mixture of categorical and quantitative variables, and the model is learned over a dataset that is supposed not to contain any anomaly. We then use the model over temporal data, potentially a data stream, using a version of the two-sided CUSUM algorithm. The proposed decision statistic is based on a conditional likelihood ratio computed for each variable given the others. Our results show that this function allows to detect anomalies variable by variable, and thus to localise the variables involved in the anomalies more precisely than univariate methods based on simple marginals.
Type de document :
Communication dans un congrès
ICML 2016 Anomaly Detection Workshop, Jun 2016, New York, United States
Liste complète des métadonnées

https://hal-imt.archives-ouvertes.fr/hal-01347167
Contributeur : Romain Laby <>
Soumis le : mercredi 20 juillet 2016 - 15:03:13
Dernière modification le : jeudi 11 janvier 2018 - 06:23:39

Fichiers

tex.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01347167, version 1
  • ARXIV : 1607.05974

Citation

Romain Laby, François Roueff, Alexandre Gramfort. Anomaly Detection and Localisation using Mixed Graphical Models. ICML 2016 Anomaly Detection Workshop, Jun 2016, New York, United States. 〈hal-01347167〉

Partager

Métriques

Consultations de la notice

253

Téléchargements de fichiers

185