cc.mallet.cluster.neighbor_evaluator
Interface NeighborEvaluator

All Known Implementing Classes:
ClassifyingNeighborEvaluator, MedoidEvaluator, PairwiseEvaluator, RandomEvaluator, RankingNeighborEvaluator

public interface NeighborEvaluator

Scores the value of changing the current Clustering to the modified Clustering specified in a Neighbor object. A common implementation of this interface uses a Classifier to assign a score to a Neighbor.

Since:
1.0
Version:
1.0
Author:
"Aron Culotta"

Method Summary
 double evaluate(Neighbor neighbor)
           
 double[] evaluate(Neighbor[] neighbors)
           
 void reset()
          Reset the state of the evaluator.
 

Method Detail

evaluate

double evaluate(Neighbor neighbor)
Parameters:
neighbor -
Returns:
A higher score indicates that the modified Clustering is preferred.

evaluate

double[] evaluate(Neighbor[] neighbors)
Parameters:
neighbors -
Returns:
One score per neighbor. A higher score indicates that the modified Clustering is preferred.

reset

void reset()
Reset the state of the evaluator.