Science Advances

This PDF file includes:

• Section S1. Deriving the linear system minimizing the Hamiltonian
• Section S2. Poisson generative model
• Section S3. Rewriting the energy
• Section S4. Ranks distributed as a multivariate Gaussian distribution
• Section S5. Bayesian SpringRank
• Section S6. Fixing c to control for sparsity
• Section S7. Comparing optimal β for predicting edge directions
• Section S8. Bitwise accuracy σ_b
• Section S9. Performance metrics
• Section S10. Parameters used for regularizing ranking methods
• Section S11. Supplementary tables
• Section S12. Supplementary figures
• Table S1. Pearson correlation coefficients between various rankings of faculty hiring networks.
• Table S2. Statistics for SpringRank applied to real-world networks.
• Fig. S1. Performance (Pearson correlation) on synthetic data.
• Fig. S2. Statistical significance testing using the null model distribution of energies.
• Fig. S3. Edge prediction accuracy over BTL for NCAA basketball data sets.
• Fig. S4. Bitwise edge direction prediction.
• Fig. S5. Edge prediction accuracy with twofold cross-validation.
• Fig. S6. Summary of SpringRank applied to computer science faculty hiring network.
• Fig. S7. Summary of SpringRank applied to history faculty hiring network.
• Fig. S8. Summary of SpringRank applied to business faculty hiring network.
• Fig. S9. Summary of SpringRank applied to Asian elephants network.
• Fig. S10. Summary of SpringRank applied to parakeet G1 network.
• Fig. S11. Summary of SpringRank applied to parakeet G2 network.
• Fig. S12. Summary of SpringRank applied to Tenpaṭṭi social support network.
• Fig. S13. Summary of SpringRank applied to Alakāpuram social support network.
• Reference (46)

Download PDF