Re:Predicting Google Product Shutdowns
by gwern on Saturday May 04, 2013 @11:32AM
(#5655197)
More like, 'a study prompted by noting serious flaws in an Arthur column and deciding to do better, while borrowing perhaps <1/10th of the final dataset from Arthur's data' (he had ~30, I wound up with 350).
[ Parent
| on HN ]
There are no comments on this comment yet.