1. Talk at AI4AM: AI for advanced materials - Madrid, Spain (May 2026)
    A generative material transformer using Wyckoff representation
  2. Invited talk at WE-Heraeus Seminar: Machine Learning for Spectroscopy - Brussels, Belgium (May 2025)
    Generative Transformer models for the dielectric function
  3. Invited talk at CECAM Machine Learning of First Principle Observables - Berlin, Germany (July 2024)
    Property predictions from limited and multi-fidelity datasets
  4. Contributed talk at the APS March Meeting 2022 - Chicago, USA (March 2022)
    Bias-imbalance in data- driven materials science: a case study on MODNet
  5. Contributed talk at the 17th ETSF Young Researchers' Meeting - Cagliari, Italy (September 2021)
    MODNet: property prediction for limited datasets and the bias-imbalance issue
  6. Invited talk at CECAM Mixed-Gen workshop (April 2021)
    Accurate and interpretable property prediction for limited materials datasets by feature selection and joint-learning
  7. Contributed talk at the APS Online March Meeting 2021 (March 2021)
    MODNet: property prediction for limited materials datasets by feature selection and joint-learning
  8. Poster presentation at the 2020 Virtual MRS Fall Meeting (November 2020)
    Machine Learning Materials Properties for Small Datasets. Symposium: Data Science and Automation to Accelerate Materials Development and Discovery

pierre-paul.debreuck [at] rub.de