Example usage for weka.classifiers.evaluation NominalPrediction NominalPrediction

Introduction

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

Prototype

public NominalPrediction(double actual, double[] distribution) 

Source Link

Document

Creates the NominalPrediction object with a default weight of 1.0.

Usage

From source file:meka.core.MLUtils.java

License:Open Source License

/** 
 * Convert to Weka (multi-target) Predictions. 
 * Note: currently only multi-label./*from  w w w  .j  a va 2 s  .c  o m*/
 * */
public static ArrayList<Prediction> toWekaPredictions(int y[], double p[]) {
    ArrayList<Prediction> predictions = new ArrayList<Prediction>();
    for (int i = 0; i < y.length; i++) {
        predictions.add(new NominalPrediction((double) y[i], new double[] { 1. - p[i], p[i] }));
    }
    return predictions;
}

---

From source file:mlflex.WekaInMemoryLearner.java

License:Open Source License

/** This method calculates the area under the curve for a set of predictions and is designed to support classification of more than two classes. This code was derived from Weka's source code.
 *
 * @param predictions Predictions that have been made
 * @return Area under the curve, weighted by the proportion of instances for each class
 * @throws Exception/*from  w  w  w. ja  va  2 s  .c  o m*/
 */
//
public static double CalculateWeightedAreaUnderRoc(Predictions predictions) throws Exception {
    ArrayList<String> uniqueActualClasses = predictions.GetUniqueActualClasses();

    if (uniqueActualClasses.size() == 0)
        return Double.NaN;

    if (predictions.Size() == 1) {
        if (predictions.Get(0).WasCorrect())
            return 1.0;

        return 0.5;
    }

    if (uniqueActualClasses.size() == 1)
        return 0.5;

    ArrayList<String> dependentVariableClasses = Utilities.ProcessorVault.DependentVariableDataProcessor
            .GetUniqueDependentVariableValues();

    FastVector predictionVector = new FastVector();
    for (Prediction prediction : predictions)
        predictionVector.addElement(
                new NominalPrediction(dependentVariableClasses.indexOf(prediction.DependentVariableValue),
                        Lists.ConvertToDoubleArray(prediction.ClassProbabilities)));

    double aucTotal = 0;

    for (int i = 0; i < dependentVariableClasses.size(); i++) {
        String dependentVariableClass = dependentVariableClasses.get(i);

        Instances result = new ThresholdCurve().getCurve(predictionVector, i);
        double auc = ThresholdCurve.getROCArea(result);

        if (!Instance.isMissingValue(auc))
            aucTotal += (auc * new PredictionResults(predictions)
                    .GetNumActualsWithDependentVariableClass(dependentVariableClass));
    }

    return aucTotal / predictions.Size();
}

---