[
https://issues.apache.org/jira/browse/SPARK-10668?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14904298#comment-14904298
]
Yanbo Liang commented on SPARK-10668:
-------------------------------------
[~mengxr] If [~lewuathe] is busy with other issues, I can take over this task.
> Use WeightedLeastSquares in LinearRegression with L2 regularization if the
> number of features is small
> ------------------------------------------------------------------------------------------------------
>
> Key: SPARK-10668
> URL: https://issues.apache.org/jira/browse/SPARK-10668
> Project: Spark
> Issue Type: New Feature
> Components: ML
> Reporter: Xiangrui Meng
> Assignee: Kai Sasaki
> Priority: Critical
>
> If the number of features is small (<=4096) and the regularization is L2, we
> should use WeightedLeastSquares to solve the problem rather than L-BFGS. The
> former requires only one pass to the data.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
