Interface RegressionEvaluation

All Superinterfaces:
Evaluation<Regressor>, com.oracle.labs.mlrg.olcut.provenance.Provenancable<EvaluationProvenance>
public interface RegressionEvaluation extends Evaluation<Regressor>
Defines methods that calculate regression performance.

  • Method Summary

    Modifier and Type
    Method
    Description
    double
    averagedExplainedVariance()
    The average explained variance across all dimensions.
    double
    averageMAE()
    The average Mean Absolute Error across all dimensions.
    double
    averageR2()
    The average R2 across all dimensions.
    double
    averageRMSE()
    The average RMSE across all dimensions.
    Map<Regressor, Double>
    explainedVariance()
    Calculatest the explained variance for all dimensions.
    double
    explainedVariance(Regressor variable)
    Calculates the explained variance of the ground truth using the predictions for the supplied dimension.
    Map<Regressor, Double>
    mae()
    Calculates the Mean Absolute Error for all dimensions.
    double
    mae(Regressor variable)
    Calculates the Mean Absolute Error for that dimension.
    Map<Regressor, Double>
    r2()
    Calculates R2 for all dimensions.
    double
    r2(Regressor variable)
    Calculates R2 for the supplied dimension.
    Map<Regressor, Double>
    rmse()
    Calculates the RMSE for all dimensions.
    double
    rmse(Regressor variable)
    Calculates the Root Mean Squared Error (i.e., the square root of the average squared errors across all data points) for the supplied dimension.

    Methods inherited from interface org.tribuo.evaluation.Evaluation

    asMap, get, getPredictions

    Methods inherited from interface com.oracle.labs.mlrg.olcut.provenance.Provenancable

    getProvenance

  • Method Details

    • averageMAE

      double averageMAE()
      The average Mean Absolute Error across all dimensions.
      Returns:
      The average Mean Absolute Error.

    • mae

      double mae(Regressor variable)
      Calculates the Mean Absolute Error for that dimension.
      Parameters:
      variable - The regression dimension to use.
      Returns:
      The Mean Absolute Error.

    • mae

      Map<Regressor, Double> mae()
      Calculates the Mean Absolute Error for all dimensions.
      Returns:
      The Mean Absolute Error.

    • averageR2

      double averageR2()
      The average R2 across all dimensions.
      Returns:
      The average R2.

    • r2

      double r2(Regressor variable)
      Calculates R2 for the supplied dimension.
      Parameters:
      variable - The regression dimension to use.
      Returns:
      The R2.

    • r2

      Map<Regressor, Double> r2()
      Calculates R2 for all dimensions.
      Returns:
      The R2.

    • averageRMSE

      double averageRMSE()
      The average RMSE across all dimensions.
      Returns:
      The average RMSE.

    • rmse

      double rmse(Regressor variable)
      Calculates the Root Mean Squared Error (i.e., the square root of the average squared errors across all data points) for the supplied dimension.
      Parameters:
      variable - The regression dimension to use.
      Returns:
      The RMSE.

    • rmse

      Map<Regressor, Double> rmse()
      Calculates the RMSE for all dimensions.
      Returns:
      The RMSE.

    • averagedExplainedVariance

      double averagedExplainedVariance()
      The average explained variance across all dimensions.
      Returns:
      The average explained variance.

    • explainedVariance

      double explainedVariance(Regressor variable)
      Calculates the explained variance of the ground truth using the predictions for the supplied dimension.
      Parameters:
      variable - The regression dimension to use.
      Returns:
      The explained variance.

    • explainedVariance

      Map<Regressor, Double> explainedVariance()
      Calculatest the explained variance for all dimensions.
      Returns:
      The explained variance.