search

krennpa / SAEforest

https://krennpa.github.io/SAEforest/
GNU General Public License v2.0
3 stars 1 forks source link

Could SAEforest package apply to binary data #1

Open lifangfiona opened 1 year ago

lifangfiona commented 1 year ago

Thanks for this wonderful package, I was wondering whether this package could apply to classification problems

krennpa commented 1 year ago

Thank you very much. The current version of the package can be only applied to continuous data/regression problems. The main reason is, that generalizations to grouped/binary data require also an adaption of the method itself, especially the part of the model that estimates the random effects. We are currently researching on methods and their implementation, but I don't know when we are ready to implement them.

If you want to contribute to the package you are very welcome!

One interesting research direction is the following contribution: Pellagatti, M, Masci, C, Ieva, F, Paganoni, AM. Generalized mixed-effects random forest: A flexible approach to predict university student dropout. Stat Anal Data Min: The ASA Data Sci Journal. 2021; 14: 241– 257. https://doi.org/10.1002/sam.11505

Best, Patrick

AlbertoImg commented 10 months ago

Hi Krennpa, Thanks for the package. I have the same question as lifangfiona. Do you have any update/news on that? Do you know any available package that could do classification? or do all the available packages have the same issue that you mentioned? Thanks in advance Best Alberto