Class MeanPredictionDistance<U>
java.lang.Object
es.uam.eps.ir.relison.links.recommendation.metrics.novdiv.MeanPredictionDistance<U>
- Type Parameters:
U
- Type of the users.
- All Implemented Interfaces:
es.uam.eps.ir.ranksys.metrics.SystemMetric<U,U>
public class MeanPredictionDistance<U>
extends java.lang.Object
implements es.uam.eps.ir.ranksys.metrics.SystemMetric<U,U>
Metric that finds the harmonic mean of the reciprocal distances between the different target user - candidate user pairs
of the recommendation.
References:
References:
- J. Sanz-Cruzado and P. Castells. Beyond Accuracy in Link Prediction. SoMePeaS 2019.
-
Field Summary
Fields Modifier and Type Field Description (package private) DistanceCalculator<U>
calculator
The distance calculator.private int
cutoff
The cutoff of the recommendationprivate Graph<U>
graph
The training network.private java.util.List<es.uam.eps.ir.ranksys.core.Recommendation<U,U>>
recs
The list of recommendations. -
Constructor Summary
Constructors Constructor Description MeanPredictionDistance(Graph<U> graph, int cutoff)
Constructor.MeanPredictionDistance(Graph<U> graph, DistanceCalculator<U> calc, int cutoff)
Constructor. -
Method Summary
-
Field Details
-
graph
The training network. -
recs
The list of recommendations. -
cutoff
private final int cutoffThe cutoff of the recommendation -
calculator
DistanceCalculator<U> calculatorThe distance calculator.
-
-
Constructor Details
-
MeanPredictionDistance
Constructor.- Parameters:
graph
- the training graph.cutoff
- recommendation cutoff.
-
MeanPredictionDistance
Constructor.- Parameters:
graph
- the training graph.calc
- distance calculator.cutoff
- recommendation cutoff.
-
-
Method Details