Package org.neo4j.gds.similarity.knn.metrics
-
Interface Summary Interface Description SimilarityComputer -
Class Summary Class Description Cosine We compute cosine similarity (normalised dot product) and turn it into a similarity metric by moving and clamping -1..1 into 0..1 using linear transformation.Euclidean Here we calculate Euclidean similarity metrics using Euclidean dictance as described in e.g.Jaccard NullCheckingNodePropertyValues Overlap Pearson Here we compute Pearson correlation coefficient and turn that into a metric. -
Enum Summary Enum Description SimilarityMetric