The objective of the internship is to test various methodologies with a view to defining a generic and optimized scheme to quantify these uncertainties. These methodologies include the expansion of polynomial chaos, reduced order modeling, multi-level Monte Carlo, and surrogate modeling (meta-models) to enable sensitivity analysis and uncertainty quantification. It will be possible to identify and classify input parameters by their sensitivity to system outputs, whatever its physical domain. These methodologies will be tested in elementary cases of Light Water Reactors (LWR) and sodium-cooled Fast Neutron Reactors (RNRNa).