[
https://issues.apache.org/jira/browse/SPARK-8986?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14623071#comment-14623071
]
Joseph K. Bradley commented on SPARK-8986:
------------------------------------------
Thanks for looking around some! I was not really thinking of anything fancy.
I was hoping existing libraries would do something like add a small constant to
the diagonal of the covariance matrix of each Gaussian. If there is no
standard to follow, we could just do that.
It'd be interesting to investigate fancier approaches in another JIRA.
> GaussianMixture should take smoothing param
> -------------------------------------------
>
> Key: SPARK-8986
> URL: https://issues.apache.org/jira/browse/SPARK-8986
> Project: Spark
> Issue Type: New Feature
> Components: MLlib
> Reporter: Joseph K. Bradley
> Original Estimate: 144h
> Remaining Estimate: 144h
>
> Gaussian mixture models should take a smoothing parameter which makes the
> algorithm robust against degenerate data or bad initializations.
> Whomever takes this JIRA should look at other libraries (sklearn, R packages,
> Weka, etc.) to see how they do smoothing and what their API looks like.
> Please summarize your findings here.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
