Package net.sf.tweety.arg.dung.learning
Class ImprovedRiveretTheoryLearner
- java.lang.Object
-
- net.sf.tweety.arg.dung.learning.ImprovedRiveretTheoryLearner
-
public class ImprovedRiveretTheoryLearner extends java.lang.Object
Implementation of the algorithm for learning (grounded) labelings from with some improvements: Riveret, RĂ©gis, and Guido Governatori. "On learning attacks in probabilistic abstract argumentation." 2016. - instead of stopping when no undecided attacks are left, we always use the full budget. While slower, greatly improves results. - added additional parameter for pruning instead of pruning at 0. approx. -10% of budget gives better results. eg budget: 1000 threshold: -100 - added fifth rule to capture self-attacking arguments- Author:
- Lars Bengel
-
-
Constructor Summary
Constructors Constructor Description ImprovedRiveretTheoryLearner(java.util.Collection<Argument> arguments, int max_cycles)
initialize learner for the given set of arguments
-
Method Summary
Modifier and Type Method Description DungTheory
learnLabelings(java.util.ArrayList<Labeling> labelings)
learn theory without pruning discarded attacksDungTheory
learnLabelings(java.util.ArrayList<Labeling> labelings, boolean prune, int threshold)
learn random labelings from the given List until no undecided attacks are left in the theory
-
-
-
Constructor Detail
-
ImprovedRiveretTheoryLearner
public ImprovedRiveretTheoryLearner(java.util.Collection<Argument> arguments, int max_cycles)
initialize learner for the given set of arguments- Parameters:
arguments
- a set of argumentsmax_cycles
- TODO description needed
-
-
Method Detail
-
learnLabelings
public DungTheory learnLabelings(java.util.ArrayList<Labeling> labelings, boolean prune, int threshold)
learn random labelings from the given List until no undecided attacks are left in the theory- Parameters:
labelings
- a list of labelingsprune
- if true, remove discarded attacks after each stepthreshold
- TODO description needed- Returns:
- the learned dung theory
-
learnLabelings
public DungTheory learnLabelings(java.util.ArrayList<Labeling> labelings)
learn theory without pruning discarded attacks- Parameters:
labelings
- a list of labelings- Returns:
- the learned dung theory
-
-