Research Article
Deep Learning Based Syndrome Diagnosis of Chronic Gastritis
Table 2
Results of model with different multilabel learning (mean ± std.).
| | Average precision ↑ | Coverage ↓ | Hamming loss ↓ | One-error ↓ | Ranking loss ↓ |
| ML-KNN | 0.754 ± 0.031 | 0.206 ± 0.017 | 0.166 ± 0.017 | 0.380 ± 0.059 | 0.173 ± 0.020 | BSVM | 0.794 ± 0.037 | 0.180 ± 0.023 | 0.166 ± 0.022 | 0.320 ± 0.065 | 0.138 ± 0.029 | Rank-SVM | 0.682 ± 0.018 | 0.255 ± 0.029 | 0.232 ± 0.014 | 0.497 ± 0.025 | 0.227 ± 0.019 | BP-MLL | 0.514 ± 0.028 | 0.395 ± 0.036 | 0.313 ± 0.010 | 0.750 ± 0.048 | 0.390 ± 0.044 | CLR | 0.784 ± 0.024 | 0.185 ± 0.023 | 0.172 ± 0.016 | 0.343 ± 0.045 | 0.143 ± 0.021 | ECC | 0.793 ± 0.021 | 0.193 ± 0.018 | 0.150 ± 0.013 | 0.277 ± 0.038 | 0.193 ± 0.023 | REKAL | 0.781 ± 0.026 | 0.209 ± 0.024 | 0.152 ± 0.012 | 0.331 ± 0.036 | 0.167 ± 0.026 | LEAD | 0.803 ± 0.019 | 0.174 ± 0.016 | 0.151 ± 0.014 | 0.304 ± 0.034 | 0.133 ± 0.015 | DBN | 0.823 ± 0.018 | 0.158 ± 0.015 | 0.139 ± 0.014 | 0.278 ± 0.028 | 0.116 ± 0.016 |
|
|