We are offering a 6-month internship at INRAE Montpellier, starting in February 2027 (flexible), on transfer learning for forecasting lake water quality. The project will build a dataset from several lakes and investigate how hybrid physics-informed deep learning models based on Universal Differential Equations (UDEs) for surface water temperature can transfer to lakes with limited data. Julia is the main programming language; prior experience is not required.
Knowledge in machine learning, statistics and differential equations is required.
Possibility to pursue a PhD (already funded) on related topics, starting in September 2027.