This might be a pretty big feature request, but right now this codebase looks like it only supports tree based xgboost models. I have a few people that want to see some in-between, decimal solutions, and booster="gblinear" might get me that.
I was thinking of adding the inputs x1, x2, x3..., squared inputs x1^2, x2^2, x3^2, cubed, etc. to fit a polynomial equation. I don't know if it will work or not, but I'll need to make a pull request to this repository if it does turn out to be a better model.
This might be a pretty big feature request, but right now this codebase looks like it only supports tree based xgboost models. I have a few people that want to see some in-between, decimal solutions, and booster="gblinear" might get me that.
I was thinking of adding the inputs x1, x2, x3..., squared inputs x1^2, x2^2, x3^2, cubed, etc. to fit a polynomial equation. I don't know if it will work or not, but I'll need to make a pull request to this repository if it does turn out to be a better model.