Skip to content

Added early_stopping_rounds

Choose a tag to compare

View all tags
@mathias-von-ottenbreit mathias-von-ottenbreit released this 01 Mar 22:00
· 179 commits to main since this release
9.3.0
0deb86c

Changes:

  • Added the constructor parameter early_stopping_rounds with a default value of 500, meaning that if the validation loss does not improve during 500 boosting steps then boosting is aborted to save time. Due to early_stopping_rounds it may make sense to try higher values of m (max number of boosting steps).
  • Updated documentation.
  • Changed default values of m for APLRRegressor and APLRClassifier and v (learning rate) for APLRClassifier.
Assets 2
Loading