pt.tumba.ngram.bayes
Class GeneralizedChoquetIntegral

java.lang.Object
  pt.tumba.ngram.bayes.GeneralizedChoquetIntegral

public class GeneralizedChoquetIntegral
extends java.lang.Object

Use Walley generalization of the Choquet lower and upper integrals to obtain the lower and upper expectations of 2-monotone capacities.

Field Summary

(package private) double[]

results

Constructor Summary

GeneralizedChoquetIntegral(TwoMonotoneCapacity tmc, DiscreteFunction df)
Calculate the lower and upper Choquet integrals using Walley's generalization, for a total variation neighborhood.

Method Summary

private void	bound_negative(TwoMonotoneCapacity tmc, DiscreteFunction df, java.util.Vector sorted_values, double[] lps, double[] ups) Obtain the lower and upper probability for the event { df(x) < sorted_value[i] }
private void	bound_positive(TwoMonotoneCapacity tmc, DiscreteFunction df, java.util.Vector sorted_values, double[] lps, double[] ups) Obtain the lower and upper probability for the event { df(x) > sorted_value[i] }
private java.util.Vector	sort_negative(DiscreteFunction df) Collect the negative values in df and sort them in decreasing order (first value is assumed zero).
private java.util.Vector	sort_positive(DiscreteFunction df) Collect the positive values in df and sort them in increasing order (first value is assumed zero).

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

results

double[] results

Constructor Detail

GeneralizedChoquetIntegral

public GeneralizedChoquetIntegral(TwoMonotoneCapacity tmc,
                                  DiscreteFunction df)

Calculate the lower and upper Choquet integrals using Walley's generalization, for a total variation neighborhood.

Method Detail

sort_positive

private java.util.Vector sort_positive(DiscreteFunction df)

Collect the positive values in df and sort them in increasing order (first value is assumed zero).

sort_negative

private java.util.Vector sort_negative(DiscreteFunction df)

Collect the negative values in df and sort them in decreasing order (first value is assumed zero).

bound_positive

private void bound_positive(TwoMonotoneCapacity tmc,
                            DiscreteFunction df,
                            java.util.Vector sorted_values,
                            double[] lps,
                            double[] ups)

Obtain the lower and upper probability for the event { df(x) > sorted_value[i] }

bound_negative

private void bound_negative(TwoMonotoneCapacity tmc,
                            DiscreteFunction df,
                            java.util.Vector sorted_values,
                            double[] lps,
                            double[] ups)

Obtain the lower and upper probability for the event { df(x) < sorted_value[i] }