Safe Reinforcement Learning for blood glucose management on T1D cohort (test)
87.8TIRRCPO
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| RCPOShield Type=Predictive Shield2026.01 | 87.8 | 3.09 | 25 | 8.51 | 0.11 | |
| CPOShield Type=Predictive Shield2026.01 | 86.92 | 3.23 | 24.2 | 5.52 | 0.17 | |
| PPOLagShield Type=Predictive Shield2026.01 | 85.81 | 3.9 | 28.69 | 4.04 | 2.39 | |
| CRPOShield Type=Predictive Shield2026.01 | 85.17 | 3.64 | 27.42 | 10.3 | 0.31 | |
| FOCOPSShield Type=Predictive Shield2026.01 | 84.03 | 4.26 | 28.45 | 6 | 1.62 | |
| TRPOLagShield Type=Predictive Shield2026.01 | 83.76 | 4.45 | 30.69 | 6.14 | 2.67 | |
| CUPShield Type=Predictive Shield2026.01 | 74.87 | 6.84 | 35.93 | 10.24 | 5.7 | |
| PCPOShield Type=Predictive Shield2026.01 | 34.86 | 21.04 | 47.55 | 6.18 | 36.89 |