Table 1 Edge prediction with BTL as a benchmark.

Edge prediction with BTL as a benchmark.. During 50 independent trials of fivefold cross-validation (250 total folds per network), columns show the percentages of instances in which SpringRank Eq. 3 and regularized SpringRank Eq. 5 with α = 2 produced probabilistic predictions with equal or higher accuracy than BTL. Distributions of accuracy improvements are shown in Fig. 3. Center columns show accuracy σa, and right columns show σL (Materials and Methods). Italics indicate where BTL outperformed SpringRank for more than 50% of tests. NCAA Basketball data sets were analyzed 1 year at a time.

Data setType% Trials higher σa versus BTL% Trials higher σL versus BTL
SpringRank+RegularizationSpringRank+Regularization
Computer science (3)Faculty hiring100.097.2100.099.6
Alakāpuram (2)Social support99.2*99.6100.0100.0
Synthetic β = 5Synthetic98.463.276.446.4
History (3)Faculty hiring97.6*96.898.898.8
NCAA Women (1998–2017) (39)Basketball94.4*87.069.151.0
Tenpaṭṭi (2)Social support88.893.6100.0100.0
Synthetic β = 1Synthetic83.265.298.498.4
NCAA Men (1985–2017) (39)Basketball76.0*62.368.552.4
Parakeet G1 (5)Animal dominance71.2*56.841.237.2
Business (3)Faculty hiring66.8*59.239.236.8
Parakeet G2 (5)Animal dominance62.051.647.647.2

*Tests that are shown in detail in Fig. 4.