Optimal estimation of nuisance paramters and models in causal inference

Inferring cause-effect relationships is often hindered by the presence of confounders. Hence, we need to build statistical models to adjust for these confounders. However, these models are not of scientific interest. This project focuses on optimal estimation of nuisance parameters indexing these models and the optimal choice of these nuisance models, both by minimizing the MSE of the causal effect.