search

imbs-hl / ranger

A Fast Implementation of Random Forests
http://imbs-hl.github.io/ranger/
770 stars 193 forks source link

GLM trees in ranger #327

Open cargomoose opened 6 years ago

cargomoose commented 6 years ago

Could you please implement glm trees in ranger, with glms used as a (multidimensional) splitting method at every node, along with the usual RF random selection of a subset of variables at each node.

mnwright commented 6 years ago

What about mobForest?

cargomoose commented 6 years ago

Hi Marvin,

Thanks for following up. Yes, I am aware of the other implementations, especially those based on the party package, but it would be nice to have it all as part of ranger and its wonderful and extensive functionality, particularly the new quantile regression capability/

best regards,

Dr. Eugene Dubossarsky Principal Trainer Presciient http://presciient.com/ 0414573322

On 26 Jun 2018, at 1:24 am, Marvin N. Wright notifications@github.com wrote:

What about mobForest https://cran.r-project.org/package=mobForest?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/imbs-hl/ranger/issues/327#issuecomment-399990987, or mute the thread https://github.com/notifications/unsubscribe-auth/AEYsmWFkTnsAI3ThTqPfBLAv_uIxBYZSks5uAQCXgaJpZM4UYdA7.

mnwright commented 6 years ago

It's a nice feature and I'd love to have it in ranger. However, I probably won't implement it myself in the near future. Sorry for this standard reply...