Some results for the adaptive Lasso

We consider the high-dimensional linear regression model, with n observations, and p > > n variables. The adaptive Lasso uses least squares loss with a weighted l_1-penalty, where the weights are prop