Tahar Nabil
PhD, AI Research Scientist @ EDF Lab.
short bio
I am currently an AI Research Scientist at EDF Lab Saclay since 2021 and was appointed an EDF Lab Expert AI Research Scientist in 2024. Prior to that I obtained a PhD in Statistical Signal Processing from Telecom Paris (2018) and worked then for three years at EDF Lab Beijing, China, still as an AI Research Scientist. My research focus is on Deep Learning for Time Series data, with an emphasis on industrial applications in the power system.
research interests
- Deep Learning for Time Series
- Time Series Foundation Models, large scale pretraining and adaptation mechanisms
- Conditional generation of long time series, application to individual electric consumption data
- Federated Learning from heterogeneous time series
- Applied AI, deep generative models (text or graph-based) for optimal industrial process design
news
| Jun 17, 2026 | Release: TS-ICL, a continuous probabilistic Time Series Foundation Model that unifies forecasting and imputation in a single zero-shot architecture. Check the repo here. |
|---|---|
| Feb 06, 2026 | Paper accepted in TMLR |
| Nov 06, 2025 | Paper accepted at the NeurIPS 2025 BERT2S Workshop (2nd CFP) |
| Jul 15, 2025 | Paper accepted at the ECML 2025 AALTD Workshop (Oral presentation) |
| Nov 14, 2024 | I was appointed Expert Research Scientist @ EDF Lab |
