Example usage for weka.classifiers.evaluation Evaluation meanAbsoluteError

Introduction

In this page you can find the example usage for weka.classifiers.evaluation Evaluation meanAbsoluteError.

Prototype

public final double meanAbsoluteError() 

Source Link

Document

Returns the mean absolute error.

Usage

From source file:cyber009.udal.functions.StatisticalAnalysis.java

/**
 * //from   w  w  w. ja  va  2 s .  c  om
 * @param classifier
 * @param trainingDataSet
 * @param unLabelSet
 * @param classTarget
 * @return 
 */
public double posteriorDistribution(Classifier classifier, Instances trainingDataSet, Instance unLabelSet,
        double classTarget) {
    double prDistribution = 0.0D;
    double classPradic = 0.0D;
    Evaluation evaluation = null;
    try {
        evaluation = new Evaluation(trainingDataSet);
        evaluation.evaluateModelOnceAndRecordPrediction(classifier, unLabelSet);
        //evaluation.pctCorrect(); // must be show for correctness  ----------------------
        classPradic = evaluation.meanAbsoluteError();
        prDistribution = classPradic * probabilityOfTargerClass(trainingDataSet, classTarget);
        //            System.out.println(classPradic);
    } catch (Exception ex) {
        Logger.getLogger(StatisticalAnalysis.class.getName()).log(Level.SEVERE, null, ex);
    }
    return prDistribution;
}

---