Table 9 Average minimum, mean, maximum, and variance, rounded to the third decimal place, of the embedding value per feature/gene (\(F=12,096\)) and per embedding dimension (\(d=50\)) across 10 iterations of JELI
From: Joint embedding–classifier learning for interpretable collaborative filtering
None | DRKG | Hetionet | PharmKG8k | PharmKG | PrimeKG | STRING | |
|---|---|---|---|---|---|---|---|
Minimum | −0.022 | −0.462 | −0.455 | −0.457 | −0.453 | 0.022 | −0.531 |
Mean | −0.004 | 0.001 | 0.006 | 0.125 | −0.003 | 0.004 | −0.017 |
Maximum | 0.014 | 0.464 | 0.462 | 0.453 | 0.457 | 0.022 | 0.512 |
Variance | 0.0 | 0.129 | 0.126 | 0.125 | 0.218 | 0.0 | 0.109 |