As an alternative to RFR, I added a Support Vector Regression model. However, SVR doesn't seem to perform better than RFR in my tests (comparing NDCG). I also abstracted away the splitting of the dataset to LTR.py.
This branch builds upon the work I've done in #10, so that PR should be merged first.
As an alternative to RFR, I added a Support Vector Regression model. However, SVR doesn't seem to perform better than RFR in my tests (comparing NDCG). I also abstracted away the splitting of the dataset to
LTR.py.This branch builds upon the work I've done in #10, so that PR should be merged first.