Research Article
Photoplethysmography Biometric Recognition Model Based on Sparse Softmax Vector and k-Nearest Neighbor
Table 5
Comparison of recognition rates of one-, two-, and three-layer feature extraction.
| Dataset | Classifier | Average recognition rate (%) | One layer | Two layers | Three layers |
| BIDMC | k-NN | 99.86 | 99.92 | 99.95 | RF | 94.42 | 96.61 | 98.07 | LDC | 84.32 | 96.38 | 98.96 | NB | 82.91 | 95.35 | 98.66 |
| MIMIC | k-NN | 96.80 | 96.84 | 97.21 | RF | 86.98 | 90.40 | 91.50 | LDC | 81.05 | 91.65 | 92.62 | NB | 93.47 | 96.57 | 97.15 |
| CapnoBase | k-NN | 99.73 | 99.78 | 99.92 | RF | 95.52 | 97.46 | 97.59 | LDC | 89.62 | 96.95 | 98.37 | NB | 96.95 | 99.54 | 99.71 |
|
|