PHC Maïmonide 2026 (France-Israel) – A Bayesian approach to constrained Gaussian and Student-t processes with applications to regression and global optimization

This 3-year project (starting in January 2026) is funded by the 2025 PHC Maïmonide call, as part of the joint research program France-Israel.

Research team

• France
  • A. F. López-Lopera (PI in France), IMAG, Univ. de Montpellier, Inria LEMON
  • G. Toulemonde, IMAG, Univ. de Montpellier, Inria LEMON
• Israel
  • A. Pakman (PI in Israel), Ben-Gurion University

Summary

This research project aims to advance Bayesian methodologies for Gaussian and Student-t processes under target constraints (e.g., inequality, monotonicity, or convexity constraints), addressing critical challenges in engineering and machine learning.

• Constrained Student-t processes (TPs) under target constraints.
• Bayesian parameter estimation methods for constrained GPs and TPs.
• Applications to regression and optimization tasks in engineering and machine learning.

Keywords. Bayesian parameter Inference, inequality constraints, MCMC, outlier-robust distributions, stochastic processes.

To acknowledge this project: “This work was funded by the 2025 PHC Maïmonide call.”