Uploaded image for project: 'Machine Learning Library'
  1. Machine Learning Library
  2. ML-389

Add common capabilities to IClassify and IRegression interfaces

    XMLWordPrintable

    Details

    • Type: Improvement
    • Status: Resolved
    • Priority: Minor
    • Resolution: Fixed
    • Affects Version/s: None
    • Fix Version/s: 7.0.0
    • Component/s: ML_Core
    • Labels:

      Description

      It is becoming clear that certain capabilities are common to all Regression or Classification algorithms and should be implemented once within ML_Core rather than within each algorithm.  The current thoughts on the core capabilities are listed below as a basis for discussion.

      Classification:

      • Accuracy Statistics:
        • Raw Accuracy
        • Power of Discrimination (PoD) = (%correct - 1/#classes) / (1-1/#classes)
        • Extended Power of Discrimination (PoDE) (%correct - %ofMostCommonClass) / (1 - %ofMostCommonClass)
        • By Class:
          • Recall (TP/(TP+FN)
          • Precision (TP /(TP + FP))
          • False Positive Rate (FP / (FP + TN))
          • Note: FP = False Positive, TP = Total Positive (# correct), TN = Total Negative (#incorrect)
        • Confusion Matrix
      • NFold Cross Validation

      Regression:

      • Accuracy Stats
        • RSquared
        • Mean Squared Error
        • Root Mean Squared Error
        • ANOVA Stats
      • NFold Cross Validation

       

        Attachments

          Activity

            People

            • Assignee:
              rdev Roger Dev
              Reporter:
              rdev Roger Dev
            • Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

              • Created:
                Updated:
                Resolved: