You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Sean Owen (JIRA)" <ji...@apache.org> on 2019/03/02 21:35:00 UTC
[jira] [Commented] (SPARK-25781) relative importance of linear
regression
[ https://issues.apache.org/jira/browse/SPARK-25781?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16782537#comment-16782537 ]
Sean Owen commented on SPARK-25781:
-----------------------------------
PS are you referring to Shapley values?
> relative importance of linear regression
> ----------------------------------------
>
> Key: SPARK-25781
> URL: https://issues.apache.org/jira/browse/SPARK-25781
> Project: Spark
> Issue Type: New Feature
> Components: MLlib
> Affects Versions: 2.3.2
> Reporter: ruxi zhang
> Priority: Minor
> Labels: features
> Attachments: v17i01.pdf
>
>
> There is an R package relaimpo that generates relative importance for linear regression features. This method utilizes sharply value regression, which will take a long time to run on big datasets. This method is quite useful for many use cases such as attribution model in marketing. It will be great if it is written in Spark with paralleled computing, which would be producing result within a much short time.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org