Table 2.
Evaluation indicators of MF combined with five classification models for the classification results of peanut seed samples.
| Models | Train-accuracy | Test-accuracy | Log loss | Hamming loss | Time(s) |
|---|---|---|---|---|---|
| RAW-XGBoost | 88.70% | 87.42% | 7151.59 | 0.125758 | 0.6133 |
| RAW-LightGBM | 89.81% | 88.48% | 6909.82 | 0.115152 | 0.1326 |
| RAW-CatBoost | 88.24% | 87.57% | 7117.05 | 0.124242 | 0.7431 |
| RAW-GBDT | 88.44% | 87.27% | 7186.13 | 0.127273 | 0.3471 |
| RAW-SEL | 91.94% | 90.75% | 6391.74 | 0.092424 | 1.3045 |
| MF-XGBoost | 96.29% | 95.91% | 5217.42 | 0.040909 | 0.5593 |
| MF-LightGBM | 96.94% | 96.36% | 5113.80 | 0.036364 | 0.1316 |
| MF-CatBoost | 97.66% | 96.51% | 5079.27 | 0.034848 | 0.9736 |
| MF-GBDT | 95.58% | 95.30% | 5355.58 | 0.04697 | 0.3071 |
| MF-SEL | 98.57% | 97.27% | 4906.57 | 0.027273 | 1.2678 |