[jira] [Updated] (SPARK-16949) Add lower/uppper bounds for AUC calculation

["Sean Owen (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/SPARK-16949?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Sean Owen updated SPARK-16949:
------------------------------
    Affects Version/s:     (was: 2.1.0)
                       2.0.0
     Target Version/s:   (was: 2.1.0)
        Fix Version/s:     (was: 2.1.0)

> Add lower/uppper bounds for AUC calculation
> -------------------------------------------
>
>                 Key: SPARK-16949
>                 URL: https://issues.apache.org/jira/browse/SPARK-16949
>             Project: Spark
>          Issue Type: Improvement
>          Components: MLlib
>    Affects Versions: 2.0.0
>            Reporter: Patrick Baier
>            Priority: Minor
>              Labels: features
>   Original Estimate: 48h
>  Remaining Estimate: 48h
>
> For some classification models only a certain part under the ROC curve is of interest.
> For instance, if I am mainly interested in avoiding false positives, I only want to compare model performance in the first section of the ROC curve (e.g. from 0 to 0.1).
> This ticket should add support for defining a lower and a upper bound (both between 0 and 1) when calling the AUC method in Mllib.
> The AUC method then only returns the AUC between these two bounds.
> The defaults of these bounds are set to 0 and 1, so there will be no break in the existing API.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org