h2o4gpu
h2o4gpu copied to clipboard
Quantile regression with h2o4gpu.gradient_boosting_regressor() not working
trafficstars
When I set loss=quantile and alpha=0.9, the result is still the same as the regression with loss=ls. I wonder if there are any other parameters to be set as well, or it is a bug.
I've created a simple PR to reproduce the issue. https://github.com/h2oai/h2o4gpu/pull/812