Subject PostDoc

This postdoctoral research project aims to explore domain adaptation approaches in the field of causal inference. We will begin by addressing the case where the graph under consideration is unknown, and then examine the case where the graph is known in one domain but partially unknown in another. The work will initially focus on the case of two domains but may be extended to multi- source scenarios (see, for example, the work of Theo Gnassou). The targeted applications are in the insurance sector. The work is in collaboration with A. Charpentier at the University of Montreal, who is exploring databases covering the various provinces of Canada. The postdoctoral position is funded by the PEPR IA and co- supervised with E. Devijver of the LIG in Grenoble. This will require frequent travel