[
https://issues.apache.org/jira/browse/SPARK-5017?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Xiangrui Meng resolved SPARK-5017.
----------------------------------
Resolution: Fixed
Fix Version/s: 1.3.0
Issue resolved by pull request 3871
[https://github.com/apache/spark/pull/3871]
> GaussianMixtureEM should use SVD for Gaussian initialization
> ------------------------------------------------------------
>
> Key: SPARK-5017
> URL: https://issues.apache.org/jira/browse/SPARK-5017
> Project: Spark
> Issue Type: Improvement
> Components: MLlib
> Affects Versions: 1.2.0
> Reporter: Joseph K. Bradley
> Assignee: Travis Galoppo
> Fix For: 1.3.0
>
>
> GaussianMixtureEM effectively does 2 matrix decompositions in Gaussian
> initialization (pinv and det). Instead, it should do SVD and use that result
> to compute the inverse and det. This will also prevent failure when the
> matrix is singular.
> Note: Breeze pinv fails when the matrix is singular:
> [https://github.com/scalanlp/breeze/issues/304]
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
