i have classification task managed train mlr
package using lda ("classif.lda") in few seconds. when trained using "classif.rpart" training never ended.
is there different setup done different methods?
my training data here if needed replicate problem. tried train with
pred.bin.task <- makeclassiftask(id="countycrime", data=dftrain, target="count.bins") train("classif.rpart", pred.bin.task)
in general, don't need change setup when switching learners -- 1 of main points of mlr
make easy! not mean it'll work though, different learning methods different things under hood.
it looks in particular case model takes long time train, didn't wait long enough complete. have quite large data frame.
looking @ data, seem have interval of values in count.bins
. treated factor r (i.e. intervals same if string matches completely), not want here. encode start , end separate (numerical) features.
Comments
Post a Comment