search

saattrupdan / doubt

Bringing back uncertainty to machine learning.
MIT License
50 stars 3 forks source link

Implement general QuantileRegressor #22

Closed saattrupdan closed 3 years ago

saattrupdan commented 3 years ago

This should wrap around any model and train it using quantile loss.

The linear case should use this wrapper if the feature matrix is singular.

saattrupdan commented 3 years ago

Implemented in https://github.com/saattrupdan/doubt/commit/c3d092ce7ccba8f33eb0dc47ab479255c9c0f11c.