Detecting masquerades with principal component analysis based on cross frequency weights

Wei Wang (1) , Sylvain Gombault (1)

1 RSM - Département Réseaux, Sécurité et Multimédia

In this paper, several cross frequency weights are used for extracting attributes of audit events. Principal Component Analysis (PCA) are then employed to discover the interrelationships and dependencies among features in a large number of variables and also to reduce the high dimensionality of these variables. Command data are used in the experiments for masquerade detection and the results demonstrate the effectiveness and efficiency of the method.

Mots clés: Intrusion detection; Principal component analysis; Masquerade detection

Domaines: Multimédia [cs.MM]; Réseaux et télécommunications [cs.NI]

Wei Wang, Sylvain Gombault. Detecting masquerades with principal component analysis based on cross frequency weights. HP-SUA 2007: 14th annual workshop of HP Software University Association, July 8-11, Munich, Germany, Jul 2007, Munich, Germany. pp.227 - 232. ⟨hal-02121414⟩