Skip to main content

Table 6 The LPbyCD parameter optimization via internal cross-validation on generic data sets

From: LPbyCD: a new scalable and interpretable approach for Link Prediction via Community Detection in bipartite networks

Data set

CD method/measure

Internal validation

External validation

AUC

\(\sigma\)

AUPR

\(\sigma\)

AUC

\(\sigma\)

AUPR

\(\sigma\)

MovieLens

Spectral part./JC\(_{CC}\)

0.82

0.00

0.22

0.02

0.82

0.01

0.26

0.02

Spectral part./JC\(_{NC}\)

0.88

0.00

0.30

0.01

0.89

0.00

0.37

0.01

Louvain/JC\(_{CC}\)

0.79

0.01

0.18

0.00

0.77

0.01

0.22

0.01

Louvain/JC\(_{NC}\)

0.88

0.00

0.26

0.00

0.88

0.00

0.34

0.01

Unicodelang

Spectral part./JC\(_{CC}\)

0.69

0.01

0.04

0.01

0.68

0.03

0.02

0.01

Spectral part./JC\(_{NC}\)

0.78

0.01

0.14

0.01

0.78

0.01

0.17

0.05

Louvain/JC\(_{CC}\)

0.73

0.00

0.09

0.01

0.72

0.02

0.10

0.03

Louvain/JC\(_{NC}\)

0.80

0.01

0.13

0.01

0.81

0.02

0.16

0.05

  1. The results are summarized in Table 6 with best values for each data set and CD method highlighted in bold