[ENH] Add LightGBMLSS to skpro

[joshdunnlime]

Is your feature request related to a problem? Please describe.
This would closely follow the work done on XGBoostLSS. Although the models are quite similar, XGBoost and LightGBM both have unique benefits such that it would be worth having both of these as an option.

Describe the solution you'd like

from skpro.regression import lightgbmlss

lgblss = lightgbmlss.LightGBMLSS()

[fkiraly]

subissue of #184 which is the umbrella for all statmixedml models

[fkiraly]

I think, a good first issue?

[joshdunnlime]

I would be interested in picking this up at some point. However I would prioritise the SupportTransform and the to_numpy/_get_bc_params_dict issues and have already made progress on these.

[fkiraly]

Sure! "good first issue" means, anyone is invited to pick this up, as it is does not require too deep understanding of the code base.

[Abdulwahid7027]

Hi, can I get this issue assigned to me? I would like to work on adding LightGBMLSS to skpro. Thanks!

[Abdulwahid7027]

If possible, I would also appreciate some guidance on the expected structure. Thanks!

[fkiraly]

@Abdulwahid7027, that would be great!

You can look at:

• extension_templates/regression.py for how the estimator should look like
• the XgboostLSS estimator for a very similar example

Please feel free to let us know on discord if you have questions!