1: $ our estimator achieves the convergence rate $