TracInAD: Measuring Influence for Anomaly Detection - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

TracInAD: Measuring Influence for Anomaly Detection

Résumé

As with many other tasks, neural networks prove very effective for anomaly detection purposes. However, very few deep-learning models are suited for detecting anomalies on tabular datasets. This paper proposes a novel methodology to flag anomalies based on TracIn, an influence measure initially introduced for explicability purposes. The proposed methods can serve to augment any unsupervised deep anomaly detection method. We test our approach using Variational Autoencoders and show that the average influence of a subsample of training points on a test point can serve as a proxy for abnormality. Our model proves to be competitive in comparison with state-of-the-art approaches: it achieves comparable or better performance in terms of detection accuracy on medical and cyber-security tabular benchmark data.

Dates et versions

hal-04244954 , version 1 (16-10-2023)

Identifiants

Citer

Hugo Thimonier, Fabrice Popineau, Arpad Rimmel, Bich-Lien Doan, Fabrice Daniel. TracInAD: Measuring Influence for Anomaly Detection. IJCNN 2022 - International Joint Conference on Neural Networks, Jul 2022, Padoue, Italy. ⟨10.1109/IJCNN55064.2022.9892058⟩. ⟨hal-04244954⟩
14 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More