Uploaded image for project: 'Spark'
  1. Spark
  2. SPARK-15699

Add chi-squared test statistic as a split quality metric for decision trees

    Details

    • Type: New Feature
    • Status: In Progress
    • Priority: Minor
    • Resolution: Unresolved
    • Affects Version/s: 2.0.0
    • Fix Version/s: None
    • Component/s: ML, MLlib
    • Labels:
      None

      Description

      Using test statistics as a measure of decision tree split quality is a useful split halting measure that can yield improved model quality. I am proposing to add the chi-squared test statistic as a new impurity option (in addition to "gini" and "entropy") for classification decision trees and ensembles.

      I wrote a blog post that explains some useful properties of test-statistics for measuring split quality, with some example results:
      http://erikerlandson.github.io/blog/2016/05/26/measuring-decision-tree-split-quality-with-test-statistic-p-values/

      (Other test statistics are also possible, for example using the Welch's t-test variant for regression trees, but they could be addressed separately)

        Issue Links

          Activity

          Hide
          apachespark Apache Spark added a comment -

          User 'erikerlandson' has created a pull request for this issue:
          https://github.com/apache/spark/pull/13440

          Show
          apachespark Apache Spark added a comment - User 'erikerlandson' has created a pull request for this issue: https://github.com/apache/spark/pull/13440
          Hide
          srowen Sean Owen added a comment -

          (if you title your PR with the JIRA number it will auto link)

          Show
          srowen Sean Owen added a comment - (if you title your PR with the JIRA number it will auto link)
          Hide
          josephkb Joseph K. Bradley added a comment -

          Erik Erlandson Just a warning: There are a lot of doc fixes for 2.0 + high priority issues, so it may take a while for this to get looked at.

          Show
          josephkb Joseph K. Bradley added a comment - Erik Erlandson Just a warning: There are a lot of doc fixes for 2.0 + high priority issues, so it may take a while for this to get looked at.

            People

            • Assignee:
              Unassigned
              Reporter:
              eje Erik Erlandson
            • Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

              • Created:
                Updated:

                Development