Postdoc – Temporal Graph Auto-Encoders for Anomaly Detection in Industrial Internet of Things

When:
15/11/2024 all-day
2024-11-15T01:00:00+01:00
2024-11-15T01:00:00+01:00

Offre en lien avec l’Action/le Réseau : – — –/– — –

Laboratoire/Entreprise : LISIC – Univ. Littoral Côte d’Opale
Durée : 24 months
Contact : esteban.bautista@univ-littoral.fr
Date limite de publication : 2024-11-15

Contexte :
The Industrial Internet of Things (IIoT) connects industrial devices, sensors, and machines to the internet, facilitating real-time data exchange. Detecting anomalies in IIoT systems is critical for ensuring security and operational efficiency. Anomalies can arise from two main data sources: communication logs, where devices are vulnerable to attacks or intrusions, and measurements, where equipment issues can disrupt production. While IIoT data can be very well modeled as an attributed temporal graph, most approaches tackle the above challenges using graph or time-series approaches, which often results in loss of important information. This postdoc project proposes to develop new algorithms that directly analyze temporal graphs, capturing both dynamic and structural information to improve anomaly detection in IIoT systems.

Sujet :
The recruited postdoc will collaborate with us on three research axes:
————————–
1. Temporal graph-based auto-encoder architecture.

We aim to build an anomaly detector based on an auto-encoder architecture that extends time-series and static graph ones to temporal graphs. We plan to address this challenge by building upon recent results that combine graph theory and signal processing into a unified formalism for temporal graphs.

————————–
2. Detection of network intrusions and attacks in communication logs.

We plan to evaluate the proposed architecture in real-world IIoT traffic datasets that capture various types of intrusions and
attacks. We would like in particular to exploit the properties of the latent space to localize the anomalous temporal graph region.

————————–
3. Detection of faults in IIoT measurements.

We aim to explore whether capturing the non-stationary dependencies of sensor measurements through temporal graphs can enhance the time-space localization of anomalies in applications such as machine health monitoring and transportation network monitoring.

Profil du candidat :
We look for highly motivated candidates with relevant experience in anomaly detection, graph machine learning, and/or deep learning. Experience in Python programming, cybersecurity and/or streaming algorithms is a plus. Ideal candidates will have a publication record in selective AI conferences.

Formation et compétences requises :
PhD in computer science or related fields.

Adresse d’emploi :
Saint-Omer, France

Document attaché : 202410062130_postdoc_ADIIOT_description_.pdf