Profile picture

Abdellah Rahmani

EPFL STI IEM LTS4
ELD 241 (Bâtiment ELD)
Station 11
1015 Lausanne

EPFLETUEDOCEDEE

Expertise

Graphs, Graph Neural network, Causal structure learning, Causal representation learning, Representation learning, Interpretability
I am a PhD candidate at EPFL under the supervision of Prof. Pascal Frossard. My research focuses on learning interpretable, meaningful causal mechanisms from data to achieve transparent, trustworthy AI. I have developed models such as CASTOR and FANTOM for uncovering causal graphs in non-stationary temporal data. Prior to joining EPFL, I studied my master’s degree at ENS Paris-Saclay (Mathematiques, Vision et Apprentissage: MVA), where I explored various aspects of AI and statistics. I also interned at Deezer Research, working on cross domain recommendation. 

Selected publications

Flow-Based Non-stationary Temporal Regime Causal Structure Learning

A. Rahmani, P. Frossard
Published in NeurIPS, 39th Conference on Neural Information Processing Systems in 2025

Causal Temporal Regime Structure Learning

A. Rahmani, P. Frossard
Published in AISTATS - The 28th International Conference on Artificial Intelligence and Statistics in 2025

A META-GNN approach to personalized seizure detection and classification

A. Rahmani, A. Venkitaraman, P. Frossard
Published in ICASSP - IEEE International Conference on Acoustics, Speech and Signal Processing in 2023