madlib-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From orhankislal <...@git.apache.org>
Subject [GitHub] madlib pull request #289: RF: Add impurity variable importance
Date Tue, 10 Jul 2018 00:39:25 GMT
GitHub user orhankislal opened a pull request:

    https://github.com/apache/madlib/pull/289

    RF: Add impurity variable importance

    JIRA: MADLIB-1205
    
    This commit makes the following changes:
    - Add impurity variable importance for random forests.
    - Rename current cat_var_importance and con_var_importance measurements to
    oob_cat_var_importance and oob_con_var_importance.
    
    New impurity measurement is provided as impurity_var_importance, and supports
    grouping. It combines the importance values for both categorical and
    continuous features into a single array.
    
    Co-authored-by: Rahul Iyer <riyer@pivotal.io>
    Co-authored-by: Jingyi Mei <jmei@pivotal.io>
    Co-authored-by: Arvind Sridhar <asridhar@pivotal.io>
    Co-authored-by: Nandish Jayaram <njayaram@apache.org>

You can merge this pull request into a Git repository by running:

    $ git pull https://github.com/madlib/madlib rf_gini_importance

Alternatively you can review and apply these changes as the patch at:

    https://github.com/apache/madlib/pull/289.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

    This closes #289
    
----
commit 622d46a85f4264fdc94bd41dc66a23f1aa2c3ed6
Author: Rahul Iyer <riyer@...>
Date:   2018-07-10T00:34:33Z

    RF: Add impurity variable importance
    
    JIRA: MADLIB-1205
    
    This commit makes the following changes:
    - Add impurity variable importance for random forests.
    - Rename current cat_var_importance and con_var_importance measurements to
    oob_cat_var_importance and oob_con_var_importance.
    
    New impurity measurement is provided as impurity_var_importance, and supports
    grouping. It combines the importance values for both categorical and
    continuous features into a single array.
    
    Co-authored-by: Rahul Iyer <riyer@pivotal.io>
    Co-authored-by: Jingyi Mei <jmei@pivotal.io>
    Co-authored-by: Arvind Sridhar <asridhar@pivotal.io>
    Co-authored-by: Nandish Jayaram <njayaram@apache.org>

----


---

Mime
View raw message