Confusion Matrix Doesn't Sum to Example #

lsevel

Hi all,

I'm running a binomial classification process. Pretty simple, retrieve data-->select attributes-->cross-validation. Somehow when I look at the results the confusion matrix adds to 26 examples when the full set has 33. When I scroll over the various operators, 33 examples are listed as moving through the process, however. I've tried the same process but selecting different attributes and do not get this problem.

Has anyone encountered this problem or have any suggestions?

Thanks!

marcin_blachnik

Maybe for some vectors you don't get results. Perhaps some values for the input training attributes are unknown (missing), your model doesn't know the answer thou you don't get it in confusion matrix

MartinLiebig

Do you have weights on your examples? The performance operator uses the weights by default.