Comments on: Eccentricity Explained http://33bits.org/2008/10/03/eccentricity-explained/ The End of Anonymized Data and What to Do About It Tue, 06 Dec 2011 23:45:54 +0000 hourly 1 http://wordpress.com/ By: Arvind http://33bits.org/2008/10/03/eccentricity-explained/#comment-9 Wed, 08 Oct 2008 03:48:03 +0000 http://33bits.wordpress.com/?p=32#comment-9 It was certainly very counterintuitive to me when I first came across it, which was long ago. It was an ‘aha’ moment for me; that’s probably when I realized the power of statistical inference and all that. It might very well have subconsciously affected my career choice :-)

Do you see a proof of optimality of this predictor, though? Why is it better than say, something that looks at the two largest samples?

]]> By: Raghu M http://33bits.org/2008/10/03/eccentricity-explained/#comment-8 Tue, 07 Oct 2008 20:11:41 +0000 http://33bits.wordpress.com/?p=32#comment-8 The puzzle is kind of counterintuitive … my first guess was that twice the average would be the best (but this only gives about 1/sqrt(n) accuracy).

]]> By: Ian Thomas http://33bits.org/2008/10/03/eccentricity-explained/#comment-7 Fri, 03 Oct 2008 17:34:59 +0000 http://33bits.wordpress.com/?p=32#comment-7 Arnand,

Thanks for the comment on my blog, which led me here. And thanks for this explanation of eccentricity – it’s some time since I’ve exercised my brain in this way!

Cheers,
Ian Thomas

]]> By: countrycousin http://33bits.org/2008/10/03/eccentricity-explained/#comment-6 Fri, 03 Oct 2008 13:23:08 +0000 http://33bits.wordpress.com/?p=32#comment-6 “counterintuitive?”

Yes. :-{

]]>