Table 6.
Performance of physicochemical descriptor classifier in a 5-fold cross-validation test.
| Cluster # | ACC (%) | ERR (%) | MCC | PER (%) | SEN (%) | SPE (%) | FN | FP | TP | TN |
|---|---|---|---|---|---|---|---|---|---|---|
| cluster-0 | 98.63 | 0.35 | 0.91918 | 96.15 | 89.29 | 99.62 | 3 | 1 | 25 | 264 |
| cluster-1 | 98.98 | 0.69 | 0.72522 | 66.67 | 80 | 99.31 | 1 | 2 | 4 | 286 |
| cluster-2 | 98.98 | 0.01 | 0.8483 | 100 | 72.73 | 100 | 3 | 0 | 8 | 282 |
| cluster-3 | 98.63 | 0.35 | 0.9356 | 97.06 | 91.67 | 99.61 | 3 | 1 | 33 | 256 |
| cluster-4 | 98.98 | 1.02 | 0.93961 | 89.29 | 100 | 98.88 | 0 | 3 | 25 | 265 |
| cluster-5 | 100 | 0 | 1 | 100 | 100 | 100 | 0 | 0 | 7 | 286 |
| cluster-6 | 100 | 0 | 1 | 100 | 100 | 100 | 0 | 0 | 23 | 270 |
| cluster-7 | 98.98 | 0.69 | 0.96881 | 96.72 | 98.33 | 99.14 | 1 | 2 | 59 | 231 |
| cluster-8 | 99.66 | 0.34 | 0.97797 | 96 | 100 | 99.63 | 0 | 1 | 24 | 268 |
| cluster-9 | 98.98 | 0.35 | 0.93086 | 95.65 | 91.67 | 99.63 | 2 | 1 | 22 | 268 |
| cluster-10 | 100 | 0 | 1 | 100 | 100 | 100 | 0 | 0 | 39 | 254 |
| cluster-11 | 99.32 | 0.01 | 0.90134 | 100 | 81.82 | 100 | 2 | 0 | 9 | 282 |
| Overall | 99.03 | 0.01 | 0.9367 | 94.20 | 94.20 | 99.47 | - | - | - | - |
ACC: accuracy, ERR: error, MCC: Matthews Correlation Coefficient, PER: Precision, SEN: Sensitivity, SPE: Specificity, FN: False Negatives, FP: False Positives, TP: True Positives, TN: True Negatives.