Table 1 Summary of PLS LOOCV and iCONIC prediction results for the human subject dataset
From: Spectroscopic approach for dynamic bioanalyte tracking with minimal concentration information
| Â | Â | Conventional PLS Model | ICONIC Model | ||||
|---|---|---|---|---|---|---|---|
Subject | No. of Data Points | RMSECV (mM) | RMSEP (mM) | k1 | k2 | λ | Change in error (%) |
1 | 25 | 1.17 | 0.68 | 0.0200 | 0.0192 | 0.89 | 42.15 |
2 | 26 | 0.72 | 0.57 | 0.0185 | 0.0178 | 0.77 | 20.39 |
3 | 26 | 1.28 | 0.53 | 0.0157 | 0.0151 | 0.67 | 58.87 |
4 | 20 | 0.76 | 0.37 | 0.0245 | 0.0240 | 0.36 | 51.31 |
5 | 32 | 0.56 | 0.46 | 0.0125 | 0.0120 | 0.55 | 17.62 |
6 | 25 | 0.80 | 0.64 | 0.0230 | 0.0217 | 0.38 | 19.47 |
7 | 26 | 0.83 | 0.44 | 0.0220 | 0.0214 | 0.40 | 46.55 |
8 | 28 | 0.79 | 0.57 | 0.0155 | 0.0148 | 0.25 | 27.84 |