[
https://issues.apache.org/jira/browse/SPARK-8078?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sean Owen updated SPARK-8078:
-----------------------------
Target Version/s: (was: 1.3.1)
Affects Version/s: (was: 1.2.2)
(was: 1.2.1)
(was: 1.3.0)
(was: 1.2.0)
(was: 1.1.1)
(was: 1.0.2)
(was: 1.0.1)
(was: 1.1.0)
Fix Version/s: (was: 1.3.1)
[~yang1111] there are a number of problems with your JIRA and PR. PLease read
https://cwiki.apache.org/confluence/display/SPARK/Contributing+to+Spark first.
For example, 1.3.1 is already released, and you should not set target version
or fix version anyway -- it's not even fixed.
You also need to explain why you think this is an important change since you're
changing behavior.
> Spark MLlib Decision Trees Improvement
> --------------------------------------
>
> Key: SPARK-8078
> URL: https://issues.apache.org/jira/browse/SPARK-8078
> Project: Spark
> Issue Type: Improvement
> Components: MLlib
> Affects Versions: 1.3.1
> Environment: ubuntu14.04
> Reporter: yangqiao
> Priority: Minor
> Labels: performance
> Original Estimate: 24h
> Remaining Estimate: 24h
>
> In Spark MLlib, Decision Trees use Gini impurity, Entropy and Variance as
> impurity. The Entropy impurity implement by calculating the Info Gain, which
> is put forward by J. Ross Quinlan in ID3 algorithm. And it can be improved by
> implementing C4.5 algorithm,which using Info Gain Ratio instead of Info Gain
> to calculate impurity. By implementing C4.5 algorithm, the Decision Trees
> model can achieve higher forecast accuracy.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
