Asymptotics / universal learning


'Maximum likelihood is efficient' 

Case-study