Semi-Supervised linear regression

We study a regression problem where for some part of the data we observe both the label variable (Y) and the predictors (X ), while for other part of the data only the predictors are given. Such a ...