Interests

• Current research (EDF R&D)

  • Tabular foundation models
  • Transformers for tabular and time series data
  • Time series forecasting and representation learning
  • Signal reconstruction and usage/appliance detection
  • Tabular models and gradient boosting models
  • Non-Intrusive Load Monitoring (NILM) / load curve disaggregation
  • In-context learning and transfer learning
  • Deep Learning for energy data

• Previous research (Ph.D. and postdoc)

  • Point processes
  • Hyperuniformity
  • Numerical integration
  • Simulation algorithms
  • Computational statistics
  • Gravitational allocation
  • Change-point detection in Hawkes processes

• Tools

  • Python, PyTorch, Hugging Face
  • Tabular foundation models: TabICL, TabPFN, TabPFN-TS
  • Representation learning models: Mantis, TimesURL
  • LightGBM, XGBoost, CatBoost, scikit-learn
  • MLflow, Hydra, uv, Poetry