indi > RE: Distribution of Test Labels
Sep 28, 2011  02:09 AM | John Muschelli
RE: Distribution of Test Labels
Two things. 1) We would like to look at the labels themselves to know exactly who had the best algorithm for prediction.  We also wanted to see how far percent accuracy changes from validation to test data sets.

That said, 2) Two sites are particularly good in these matters.  http://kaggle.com/ and http://tunedit.org/ (the second may be easier to set up), if people wanted.

Threaded View

TitleAuthorDate
Paolo Avesani Sep 15, 2011
RE: Distribution of Test Labels
John Muschelli Sep 28, 2011
Paolo Avesani Sep 26, 2011
Satrajit Ghosh Sep 27, 2011
Maarten Mennes Sep 27, 2011
Arno Klein Sep 27, 2011
Huiguang He Sep 17, 2011
Maarten Mennes Sep 26, 2011