Research Article
TCM Prescription Generation via Knowledge Source Guidance Network Combined with Herbal Candidate Mechanism
Table 4
Performance comparison of different combinations of hyperparameters in TKMM.
| (λ1,λ2) | Precision | Recall | F1-score |
| (0.0,1.0) | 0.3396 | 0.2065 | 0.2568 | (0.1,0.9) | 0.3487 | 0.2298 | 0.2770 | (0.2,0.8) | 0.3521 | 0.2330 | 0.2804 | (0.3,0.7) | 0.3653 | 0.2509 | 0.2975 | (0.4,0.6) | 0.3739 | 0.2504 | 0.2999 | (0.5,0.5) | 0.3711 | 0.2485 | 0.2977 | (0.6,0.4) | 0.3662 | 0.2409 | 0.2906 | (0.7,0.3) | 0.3619 | 0.2337 | 0.2840 | (0.8,0.2) | 0.3550 | 0.2291 | 0.2785 | (0.9,0.1) | 0.3498 | 0.2134 | 0.2651 | (1.0,0.0) | 0.3402 | 0.1968 | 0.2493 |
|
|