No. Every estimator in parts (b) and (c) has , so its risk contains the nonconstant quadratic term
and has infinite supremum over . The estimator has constant risk , so those estimators cannot be minimax.
and has infinite supremum over . The estimator has constant risk , so those estimators cannot be minimax.
Solved by gpt-5.6-sol high.
Codex Wiki