Comments on: ABC as knn… http://xianblog.wordpress.com/2012/09/13/abc-as-knn/ an attempt at bloggin, from scratch... Wed, 19 Jun 2013 14:52:07 +0000 hourly 1 http://wordpress.com/ By: Keith O'Rourke http://xianblog.wordpress.com/2012/09/13/abc-as-knn/comment-page-1/#comment-24663 Fri, 14 Sep 2012 19:31:36 +0000 http://xianblog.wordpress.com/?p=17513#comment-24663 Neat, if you can efficiently search Andrew G’s blog, I proposed Bayes as inherently being a k-nearest neighbour procedure in 2005/6. I had, in a pinch, come up with (what Don Rubin used in his 1984 paper) a two stage sampling model in order to explain Bayes to Epi students.

Not that I did anything with it (or could I with computational resources back then), but I remember the clear lack of enthusiam to see it that way.

It just seemed wrong or worse too simple!

]]>