Research Article
Transferability of a Machine Learning-Based Model of Hourly Traffic Volume Estimation—Florida and New Hampshire Case Study
Table 2
The detailed results for different approaches.
| | R2 | MAPE (%) | EMFR (%) | Base | Fl | TL | Ext | Base | Fl | TL | Ext | Base | Fl | TL | Ext |
| Mean | 0.72 | 0.34 | 0.77 | 0.81 | 43.3 | 61.0 | 34.4 | 34.9 | 8.12 | 12.61 | 7.36 | 7.15 | Std | 0.49 | 1.30 | 0.30 | 0.13 | 47.6 | 43.8 | 26.1 | 29.0 | 4.03 | 7.52 | 2.75 | 2.23 | min | −2.46 | −8.31 | −1.08 | 0.06 | 16.2 | 27.9 | 14.1 | 16.0 | 4.44 | 7.52 | 4.27 | 4.67 | 25 per. | 0.76 | 0.54 | 0.76 | 0.78 | 22.7 | 40.8 | 20.8 | 21.4 | 5.85 | 8.20 | 5.93 | 5.79 | 50 per. | 0.82 | 0.70 | 0.85 | 0.84 | 26.9 | 49.0 | 26.9 | 25.5 | 7.04 | 9.95 | 6.67 | 6.68 | 75 per. | 0.88 | 0.80 | 0.89 | 0.88 | 37.6 | 59.6 | 34.0 | 34.2 | 8.40 | 13.96 | 7.90 | 7.69 | max | 0.94 | 0.92 | 0.95 | 0.94 | 311 | 286 | 149 | 182 | 28.3 | 48.0 | 20.8 | 20.0 |
|
|
Bold values indicate mean and median values.
|