Uploaded image for project: 'Spark'
  1. Spark
  2. SPARK-9478

Add sample weights to Random Forest

    XMLWordPrintableJSON

Details

    • Improvement
    • Status: Resolved
    • Major
    • Resolution: Fixed
    • 1.4.1
    • 3.0.0
    • ML
    • None

    Description

      Currently, this implementation of random forest does not support sample (instance) weights. Weights are important when there is imbalanced training data or the evaluation metric of a classifier is imbalanced (e.g. true positive rate at some false positive threshold). Sample weights generalize class weights, so this could be used to add class weights later on.

      Attachments

        Issue Links

          duplicates

          New Feature - A new feature of the product, which has yet to be developed. SPARK-19591 Add sample weights to decision trees

          • Major - Major loss of function.
          • Resolved
          is blocked by

          New Feature - A new feature of the product, which has yet to be developed. SPARK-19591 Add sample weights to decision trees

          • Major - Major loss of function.
          • Resolved

          Improvement - An improvement or enhancement to an existing feature or task. SPARK-14599 BaggedPoint should support weighted instances.

          • Major - Major loss of function.
          • Resolved
          Is contained by

          Umbrella - An overarching type made of sub-tasks SPARK-14046 RandomForest improvement umbrella

          • Major - Major loss of function.
          • Resolved
          relates to

          Umbrella - An overarching type made of sub-tasks SPARK-9610 Class and instance weighting for ML

          • Major - Major loss of function.
          • Resolved
          links to

          Pull request #9008 [Github] Pull Request #9008 (rotationsymmetry)

          Pull request #13851 [Github] Pull Request #13851 (n-triple-a)

          Pull request #16722 [Github] Pull Request #16722 (sethah)

          Web Link GitHub Pull Request #27097

          Activity

            People

              podongfeng Ruifeng Zheng
              pcrenshaw Patrick Crenshaw
              Votes:
              15 Vote for this issue
              Watchers:
              22 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: