Table 4.
Dataset D: classification performance of Ph-CNN compared to other classifiers on Healthy vs. CDf classification task
| CDf | Ph-CNN | LSVM | ||||
|---|---|---|---|---|---|---|
| p | MCC | min CI | max CI | MCC | min CI | max CI |
| 65 | 0.785 | 0.775 | 0.795 | 0.781 | 0.776 | 0.785 |
| 130 | 0.832 | 0.825 | 0.840 | 0.833 | 0.829 | 0.838 |
| 195 | 0.896 | 0.891 | 0.901 | 0.910 | 0.907 | 0.912 |
| 259 | 0.927 | 0.924 | 0.930 | 0.920 | 0.918 | 0.923 |
| MLPNN | RF | |||||
| p | MCC | min CI | max CI | MCC | min CI | max CI |
| 65 | 0.604 | 0.593 | 0.614 | 0.764 | 0.760 | 0.769 |
| 130 | 0.821 | 0.817 | 0.825 | 0.805 | 0.800 | 0.810 |
| 195 | 0.830 | 0.825 | 0.836 | 0.863 | 0.860 | 0.867 |
| 259 | 0.858 | 0.854 | 0.862 | 0.880 | 0.877 | 0.883 |
The performance measure is MCC, with 95% studentized bootstrap confidence intervals (min CI, max CI). Models are computed for p={25%,50%,75% and 100%} of total number of features for each task. Comparing algorithms are linear Support Vector Machines (LSVM), Random Forest (RF) and MultiLayer Perceptron (MLPNN)