[
https://issues.apache.org/jira/browse/SPARK-7210?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14622737#comment-14622737
]
Joseph K. Bradley commented on SPARK-7210:
------------------------------------------
More thoughts from Reza: We should consider degenerate cases, and to say we
handle them "correctly," we can compare with R as a reasonable gold standard.
E.g., how does it handle normal PDFs when the covariance matrix is not full
rank?
Relatedly, we should add a smoothing parameter to GaussianMixture. That might
actually be higher priority than this JIRA. I'll make a JIRA for that and link
it from the umbrella.
> Test matrix decompositions for speed vs. numerical stability for Gaussians
> --------------------------------------------------------------------------
>
> Key: SPARK-7210
> URL: https://issues.apache.org/jira/browse/SPARK-7210
> Project: Spark
> Issue Type: Improvement
> Components: MLlib
> Reporter: Joseph K. Bradley
> Priority: Minor
>
> We currently use SVD for inverting the Gaussian's covariance matrix and
> computing the determinant. SVD is numerically stable but slow. We could
> experiment with Cholesky, etc. to figure out a better option, or a better
> option for certain settings.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
