The incoming student will work on using dimensionality reduction techniques to develop surrogate models for expensive numerical calculations (see https://www.frontiersin.org/articles/10.3389/fphy.2022.1054524/full, and https://journals.aps.org/prc/abstract/10.1103/PhysRevC.109.044612). The approach will follow a series of recent data-driven approaches (see https://arxiv.org/abs/2502.03680, https://arxiv.org/abs/2406.04279, and https://youtu.be/XiRFiH-7MG8?si=rcyS-sU4eZVEgjWa&t=1321). If the student comes with a specific model/simulation in mind the project will focus on creating an emulator for that, but other options are available ranging from reactions, nuclear structure calculations, nuclear astrophysics, and experimental control. We expect that after the end of the program the student will be able to bring to their home institution the knowledge to apply dimensionality reduction techniques to speed up computations and tailor them to their specific needs. If progress during the three month period was effective, we also expect to write a publication with the findings. Expected incoming student skills: Intermediate coding experience in at least one language (python, Mathematica, c++, fortran). Basic knowledge of differential equations. The registration for the internship must be done on the website: https://npa.in2p3.fr/summer-training-program/