Uploaded image for project: 'Spark'
  1. Spark
  2. SPARK-20348

Support squared hinge loss (L2 loss) for LinearSVC

    XMLWordPrintableJSON

Details

    • New Feature
    • Status: Resolved
    • Minor
    • Resolution: Duplicate
    • 2.2.0
    • None
    • ML
    • None

    Description

      While Hinge loss is the standard loss function for linear SVM, Squared hinge loss (a.k.a. L2 loss) is also popular in practice. L2-SVM is differentiable and imposes a bigger (quadratic vs. linear) loss for points which violate the margin. Some introduction can be found from http://mccormickml.com/2015/01/06/what-is-an-l2-svm/

      Liblinear and scikit learn both offer squared hinge loss as the default loss function for linear SVM.

      Attachments

        Issue Links

          is related to

          Improvement - An improvement or enhancement to an existing feature or task. SPARK-20602 Adding LBFGS optimizer and Squared_hinge loss for LinearSVC

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

          Sub-task - The sub-task of the issue SPARK-20503 ML 2.2 QA: API: Python API coverage

          • Blocker - Blocks development and/or testing work, production could not run
          • Resolved
          links to

          Pull request #17645 [Github] Pull Request #17645 (hhbyyh)

          Activity

            People

              Unassigned Unassigned
              yuhaoyan yuhao yang
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: