1 paper · 1 filter
Xavier D'Haultfoeuille, Christophe Gaillac, Arnaud Maurel
We study linear regressions in a context where the outcome of interest and some of the covariates are observed in two different datasets that cannot be matched. Traditional approac…