You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Joseph K. Bradley (JIRA)" <ji...@apache.org> on 2015/06/20 22:03:00 UTC
[jira] [Resolved] (SPARK-8468) Cross-validation with
RegressionEvaluator prefers higher RMSE
[ https://issues.apache.org/jira/browse/SPARK-8468?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Joseph K. Bradley resolved SPARK-8468.
--------------------------------------
Resolution: Fixed
Fix Version/s: 1.4.1
1.5.0
Issue resolved by pull request 6905
[https://github.com/apache/spark/pull/6905]
> Cross-validation with RegressionEvaluator prefers higher RMSE
> -------------------------------------------------------------
>
> Key: SPARK-8468
> URL: https://issues.apache.org/jira/browse/SPARK-8468
> Project: Spark
> Issue Type: Bug
> Components: ML
> Affects Versions: 1.4.0
> Reporter: Chelsea Zhang
> Assignee: Liang-Chi Hsieh
> Priority: Blocker
> Fix For: 1.5.0, 1.4.1
>
>
> Please correct me if I'm wrong, but RegressionEvaluator seems to implement the evaluate() method backwards. The interface expects higher return values from evaluate() to indicate better models. RegressionEvaluator returns RMSE by default - a value we should try to minimize.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org