Open mzed opened 4 years ago
If knn and nn knew which output they were looking at, there wouldn't need to be multiple copies of the training data set.
The design idea was that these models didn't need to have any data about the set they are in. Does this need to break? Or, is there another way?
This is in progress in the dev branch.
If knn and nn knew which output they were looking at, there wouldn't need to be multiple copies of the training data set.
The design idea was that these models didn't need to have any data about the set they are in. Does this need to break? Or, is there another way?