Table 8 The average ± standard deviation validation metric on the weakly correlated testing subset across \(N=100\) iterations for the Top-10 algorithms on the “Synthetic” dataset in Table 1. Average (resp., standard deviation) values are rounded up to the closest second (resp., first) decimal place.
From: Comprehensive evaluation of pure and hybrid collaborative filtering in drug repurposing
Model | NS AUC | AUC |
|---|---|---|
HAN | \(1.00 \pm 0.0\) | \(1.00 \pm 0.0\) |
Fast.ai | \(1.00 \pm 0.0\) | \(1.00 \pm 9.10^{-4}\) |
LogisticMF | \(0.99 \pm 9.10^{-4}\) | \(0.99 \pm 1.10^{-4}\) |
BNNR | \(0.76 \pm 3.10^{-3}\) | \(0.98 \pm 2.10^{-4}\) |
NIMCGCN | \(0.54 \pm 3.10^{-4}\) | \(0.97 \pm 4.10^{-6}\) |
ALSWR | \(0.50 \pm 0.0\) | – |
LibMF | \(0.45 \pm 1.10^{-16}\) | \(0.98 \pm 3.10^{-16}\) |
SCPMF | \(0.44 \pm 7.10^{-2}\) | \(0.40 \pm 1.10^{-1}\) |
LRSSL | – | \(0.15 \pm 8.10^{-3}\) |
PMF | – | \(0.08 \pm 7.10^{-3}\) |
