[
https://issues.apache.org/jira/browse/SPARK-6004?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14337791#comment-14337791
]
Liang-Chi Hsieh commented on SPARK-6004:
----------------------------------------
Okay. At least we can modify it to return best model if we do not stop early.
Another question is do we want to also return the best model (regarding
training data) even the validation is not used (validate is false) instead of
full model?
> Pick the best model when training GradientBoostedTrees with validation
> ----------------------------------------------------------------------
>
> Key: SPARK-6004
> URL: https://issues.apache.org/jira/browse/SPARK-6004
> Project: Spark
> Issue Type: Improvement
> Components: MLlib
> Reporter: Liang-Chi Hsieh
> Priority: Minor
>
> Since the validation error does not change monotonically, in practice, it
> should be proper to pick the best model when training GradientBoostedTrees
> with validation instead of stopping it early.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
