Sofiane ENNADIR

About Me

I am a PhD student in the EECS School at the KTH Royal Institute of Technology in Sweden, working under the supervision of Prof. Michalis Vazirgiannis and Prof. Henrik Boström. I am also fortunate to be part of the WASP AI program and my research is supported by the Wallenberg AI, Autonomous Systems and Software Program funded by the Knut and Alice Wallenberg Foundation. My research revolves around Graph Neural Networks (GNNs), delving deeply into their robustness and practical applications. Specifically, my PhD focuses on enhancing the robustness of these models against adversarial attacks while analyzing and developing various attack strategies. I am particularly interested in theoretically understanding the robustness behaviour of GNNs and consequently proposing efficient and viable defense methods. I have been also lately interested in other application of GNNs, such as temporal graphs.

I am currently interning at King AI Labs (Microsoft - ABK) working on Continuous-Time Dynamic Graphs (CTDG). Before that I was a summer intern at the Flatiron Institute within the Foundation Models for Science initiative, where I worked with Leopoldo Sarra and Siavash Golkar. The internship evolved around investigating the possible extension of the Joint Embeddings approach for building Time Series foundation models.

Prior to joining KTH, I completed a MSc Degree in "Applied Mathematics - Data Sciences" from l’Ecole Polytechnique in Paris, France. I also hold a Master of Engineering from EMINES, School of Industrial Management at the Mohammed VI Polytechnique University (UM6P) in Morocco.

My academic CV can be accessed here.

News

[Sep. 2024] Our paper "Joint Embedding go Temporal" is accepted to the "Time Series in the Age of Large Models" Workshop at Neurips 2024!
[Sep. 2024] Our paper "If You Want to Be Robust, Be Wary of Initialization" is accepted to Neurips 2024!
[April. 2024] Presented our GNN robustness on the Deep Learning: Classics and Trends reading group (Collective ML) [Slides].
[Mar. 2024] We presented our ICLR paper at the Morocco AI Webinar [Slides | Recording].
[Jan. 2024] Our paper "Bounding the Expected Robustness of Graph Neural Networks Subject to Node Feature Attacks" is accepted to ICLR 2024!
[Dec. 2023] Our paper "Simple and Yet Fairly Effective Defense for Graph Neural Networks" is accepted to AAAI 2024!
[Nov. 2023] Presented our paper "UnboundAttack : Generating Unbounded Adversarial Attacks to Graph Neural Networks" at 12th Internation Conference on Complex Networks. - [Slides]
[Sep. 2023] Presented "Conformalized Adversarial Attack Detection for Graph Neural Networks" at 12th Symposium on Conformal and Probabilistic Prediction with Applications. - [Slides]
[Jun. 2023] Our paper "Simple and Yet Fairly Effective Defense for Graph Neural Networks" is accepted to the AdvML workshop at ICML 2023!
[Apr. 2023] New Preprint of our work on detecting ARGs using both Sequence and Structures. This work has also been presented at the AI forScience at Neurips 2022.

Publications

^* is used to denote equal contribution.

If You Want to Be Robust, Be Wary of Initialization

S. Ennadir, J. Lutzeyer, M. Vazirgiannis & E. Bergou.

The Thirty-Eighth Annual Conference on Neural Information Processing Systems (Neurips), 2024.

PDF Code BibTex
Joint Embedding go Temporal

S. Ennadir, S. Golkar, L. Sarra.

Time Series in the Age of Large Models Workshop, Neurips 2024.

PDF Code
Bounding the Expected Robustness of Graph Neural Networks Subject to Node Feature Attacks

Y. Abbahadou^*, S. Ennadir^*, J. Lutzeyer, M. Vazirgiannis & H. Boström.

The Twelfth International Conference on Learning Representations (ICLR), 2024.

PDF Code BibTex

A Simple and Yet Fairly Effective Defense for Graph Neural Networks

S. Ennadir , Y. Abbahadou, J. Lutzeyer, M. Vazirgiannis & H. Boström.

Thirty-Eighth AAAI Conference on Artificial Intelligence (AAAI), 2024.

PDF Code BibTex
[Previous version] presented at the ICML 2nd AdvML workshop 2023.
Interpretable Graph Neural Networks for Tabular Data

A. Alkhatib, S. Ennadir, H. Boström & M. Vazirgiannis.

The 27th European Conference on Artificial Intelligence (ECAI), 2024.

PDF Code
[Previous version] presented at the DMLR workshop in ICLR 2024.

UnboundAttack : Generating Unbounded Adversarial Attacks to Graph Neural Networks

S. Ennadir, A. Alkhatib, G. Nikolentzos, M. Vazirgiannis & H. Boström.

12th International Conference on Complex Networks (CNA), 2023.

PDF Slides
Conformalized Adversarial Attack Detection for Graph Neural Networks

S. Ennadir, A. Alkhatib, H. Boström & M. Vazirgiannis.

12th Symposium on Conformal and Probabilistic Prediction with Applications (COPA), 2023.

PDF Slides BibTex
Structure-Aware Antibiotic Resistance Classification using Graph Neural Networks

A. Qabel, S. Ennadir, G. Nikolentzos, J. Lutzeyer, M. Chatzianastasis, H. Boström & M. Vazirgiannis.

AI4Science Workshop at Neurips, 2022.

PDF Code BibTex
[Extended version] currently under review.

Generating Graph Perturbations to Enhance the Generalization of GNNs

S. Ennadir, G. Nikolentzos, H. Boström & M. Vazirgiannis.

Under Review.

PDF

I would like to thank Johannes Lutzeyer for allowing me to copy the format of his website. This website is powered by Jekyll and Minimal Light theme.