View Single Post
  #6  
Old 07-21-2012, 01:57 PM
rakhlin rakhlin is offline
Member
 
Join Date: Jun 2012
Posts: 24
Default Re: Hoeffding’s inequality

I got all numbers, but don't understand the problem.

First, c_rand and c_min aren't certain coins. They're many random different coins so we can't derive their features from the experiment. Formally we can speak of c_1 only.

Second, all three nu's (their averages) converge to theoretical limits - not sure I can post actual numbers - two equal numbers and another one. In this sense all three "coins" satisfy Hoeffding’s inequality, that is they generate data that probabilistically converge to certain limits (not 0.5 in case of c_min). But we can't select all three as an answer.

So I would be grateful for clarification to this problem
Reply With Quote