[
https://issues.apache.org/jira/browse/SPARK-13712?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15193815#comment-15193815
]
Joseph K. Bradley commented on SPARK-13712:
-------------------------------------------
I don't think this is a high-priority item to add. It is much more expensive
that one-vs-rest, and there are better methods we should add first.
(Specifically, we should add a method based on error-correcting codes since it
will be cheaper and have guarantees essentially as good as one-vs-one.) I'll
comment on the PR as well. If you'd like, feel free to submit this as a Spark
package since I'm sure some people would find it useful. Thanks! I'll close
this JIRA for now.
> Add OneVsOne to ML
> ------------------
>
> Key: SPARK-13712
> URL: https://issues.apache.org/jira/browse/SPARK-13712
> Project: Spark
> Issue Type: New Feature
> Components: ML
> Reporter: zhengruifeng
> Priority: Minor
>
> Another Meta method for multi-class classification.
> Most classification algorithms were designed for balanced data.
> The OneVsRest method will generate K models on imbalanced data.
> The OneVsOne will train K*(K-1)/2 models on balanced data.
> OneVsOne is less sensitive to the problems of imbalanced datasets, and can
> usually result in higher precision.
> But it is much more computationally expensive, although each model are
> trained on a much smaller dataset. (2/K of total)
> The OneVsOne is implemented in the way OneVsRest did:
> val classifier = new LogisticRegression()
> val ovo = new OneVsOne()
> ovo.setClassifier(classifier)
> val ovoModel = ovo.fit(data)
> val predictions = ovoModel.transform(data)
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
