Joon Young Kim, I agree that $ \hat a_{ML} $ should be $ \frac{|x_i|}{2} $.

Thus $ \hat a_{ML} = {max}_i\frac{|x_i|}{2} $

I don't know how to (mathematically) show whether it is Biased or Unbiased.

Alumni Liaison

Correspondence Chess Grandmaster and Purdue Alumni

Prof. Dan Fleetwood