[
https://issues.apache.org/jira/browse/SPARK-8536?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14617638#comment-14617638
]
Joseph K. Bradley commented on SPARK-8536:
------------------------------------------
If it's easy to add both, I think we still should. It is very difficult to
evaluate LDA. But if it's much simpler to only support one type of asymmetric
prior, then that seems OK.
> Generalize LDA to asymmetric priors
> -----------------------------------
>
> Key: SPARK-8536
> URL: https://issues.apache.org/jira/browse/SPARK-8536
> Project: Spark
> Issue Type: Improvement
> Components: MLlib
> Reporter: Joseph K. Bradley
> Original Estimate: 216h
> Remaining Estimate: 216h
>
> Several users have requested LDA using asymmetric priors, including learning
> those hyperparameters. We should add this functionality.
> Note that it could be added to EM, online LDA, or both. I suggest adding it
> to online LDA initially, followed by EM in a separate PR.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
