Split into train/test folds (this would be a trivial case to split into multiple folds, possibly consider that sometime in the future as an introduction to multiclass classification or cross validation).
I made machines 13, 14, and 15 the test examples, meaning one example each of sendback, ok, fix exists in the test data.
This adds a "machines" dataset based closely on the one described in the ACE Data Mining User Guide.
not_replaceabletoirreplaceableI made machines 13, 14, and 15 the test examples, meaning one example each of
sendback,ok,fixexists in the test data.