You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@spark.apache.org by "Joseph K. Bradley (JIRA)" <ji...@apache.org> on 2015/05/07 01:27:00 UTC

[jira] [Created] (SPARK-7409) Designing multilabel abstractions for spark.ml

Joseph K. Bradley created SPARK-7409:
----------------------------------------

             Summary: Designing multilabel abstractions for spark.ml
                 Key: SPARK-7409
                 URL: https://issues.apache.org/jira/browse/SPARK-7409
             Project: Spark
          Issue Type: Brainstorming
          Components: ML
            Reporter: Joseph K. Bradley


This JIRA is for discussing how to support multi-label prediction in the Pipelines API (spark.ml package).  Some issues to figure out are:
* Should there be abstractions?
** How should they relate to the existing single-label abstractions: Predictor, Classifier, Regressor?
** How much code sharing can the abstractions provide?
* How should we support a mix of categorical and real-valued labels?
* How do we support structure among the labels?  There could be no known structure, a graphical structure, a chain structure, etc., depending on the application/model.



--
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