Re:Interview with Scott Aaronson
by ced on Saturday April 23, 2016 @12:53PM
(#11556301)
Clever solution. Is there a more general principle? I thought that most MNAR problems were hopeless, but maybe not.
[ Parent
| on HN ]
It helps that we have a rough model of the publication process that we can correct for. It's kinda like assessing the underlying distribution with only access to the tail of it, which isn't ideal, but can be done.