Package com.graphhopper.jsprit.core.algorithm.recreate

Class AverageRegretKScoringFunction

java.lang.Object

com.graphhopper.jsprit.core.algorithm.recreate.AverageRegretKScoringFunction

All Implemented Interfaces:

RegretKScoringFunction

public class AverageRegretKScoringFunction
extends Object
implements RegretKScoringFunction

Average-based regret-k scoring function.

Computes: score = mean(c_2..c_k) - c_1

This variant uses the average regret - the difference between the mean of alternatives 2 through k and the best. This provides a balanced view.

Constructor Summary

Constructors

Constructor	Description
AverageRegretKScoringFunction(ScoringFunction additionalScorer, int k)	Creates an average regret-k scoring function.

Method Summary

Modifier and Type	Method	Description
int	getK()
double	score(RegretKAlternatives alternatives, Job job)	Computes a score for a job based on its k-best insertion alternatives.

Methods inherited from class java.lang.Object

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

Constructor Details

AverageRegretKScoringFunction

public AverageRegretKScoringFunction(ScoringFunction additionalScorer,
 int k)

Creates an average regret-k scoring function.

Parameters:

additionalScorer - additional scoring function for tie-breaking (can be null)

k - the number of alternatives to consider (use Integer.MAX_VALUE for "all")

Method Details

score

public double score(RegretKAlternatives alternatives,
 Job job)

Description copied from interface: RegretKScoringFunction

Computes a score for a job based on its k-best insertion alternatives. Higher scores indicate higher urgency to insert the job now.

Specified by:

score in interface RegretKScoringFunction

Parameters:

alternatives - the k-best insertion alternatives for the job

job - the job being scored

Returns:

the regret score (higher = more urgent to insert)

getK

public int getK()