Table 4 Model performance: VNS, RPS mapping algorithms
Van Hout Crosswalk | Alava Crosswalk | |||||
|---|---|---|---|---|---|---|
M4 (Linear) | M5 (Non-Linear) | M6 (Polynomial) | M4 (Linear) | M5 (Non-Linear) | M6 (Polynomial) | |
VNS | ||||||
MAE | 0.108 | 0.019 | 0.0012 | 0.100 | 0.118 | 0.0011 |
RMSE | 0.152 | 0.021 | 0.00021 | 0.151 | 0.016 | 0.0022 |
Predicted Mean (SE) | 0.9115 (0.0003) | 0.8851 (0.0005) | 0.9081 (0.0003) | 0.9000 (0.0002) | 0.8996 (0.0003) | 0.8939 (0.0003) |
Observed Mean (SE) | 0.9067 (0.005) | 0.9067 (0.005) | 0.9067 (0.005) | 0.9008 (0.005) | 0.9008 (0.005) | 0.9008 (0.005) |
Difference (Mean) | -0.0048 | 0.0216 | -0.0014 | 0.0008 | 0.0012 | 0.0069 |
AIC | -681 | -691 | -2850 | -701 | -714 | -2899 |
Predicted QALY | 1.5967 (0.0013) | 1.5503 (0.0020) | 1.5899 (0.0013) | 1.5758 (0.0007) | 1.5742 (0.0011) | 1.5646 (0.0011) |
Observed QALY | 1.5979 (0.017) | 1.5979 (0.017) | 1.5979 (0.017) | 1.5887 (0.017) | 1.5887 (0.017) | 1.5887 (0.017) |
Difference (QALY) | 0.0012 | 0.0476 | 0.0080 | 0.0129 | 0.0145 | 0.0241 |
RPS | ||||||
MAE | 0.107 | 0.031 | 0.001 | 0.100 | 0.014 | 0.0008 |
RMSE | 0.153 | 0.031 | 0.00018 | 0.151 | 0.016 | 0.0011 |
Predicted Mean (SE) | 0.9068 (0.0004) | 0.8834 (0.0002) | 0.9069 (0.0004) | 0.8997 (0.0003) | 0.8917 (0.0003) | 0.8984 (0.0004) |
Observed Mean (SE) | 0.9067 (0.005) | 0.9067 (0.005) | 0.9067 (0.005) | 0.9008 (0.005) | 0.9008 (0.005) | 0.9008 (0.005) |
Difference (Mean) | -0.0001 | 0.0233 | -0.0002 | 0.0011 | 0.0091 | 0.0024 |
AIC | -664 | -697 | -2743 | -683 | -698.4 | -2888 |
Predicted QALY | 1.5862 (0.0014) | 1.5449 (0.0007) | 1.5859 (0.0013) | 1.5739 (0.0012) | 1.5594 (0.0011) | 1.5699 (0.0017) |
Observed QALY | 1.5979 (0.017) | 1.5979 (0.017) | 1.5979 (0.017) | 1.5887 (0.017) | 1.5887 (0.017) | 1.5887 (0.017) |
Difference (QALY) | 0.0117 | 0.053 | 0.012 | 0.0148 | 0.0293 | 0.0188 |