Table 1.
Classification performance on generated datasets. Summary statistics of the classification accuracy of the SVM (as in the first step of COMBI) and the DNN (as in the first step of DeepCOMBI) are presented. Values corresponding to accuracy and balanced accuracy in parentheses are given
Mean accuracy (balanced accuracy) | Standard deviation of accuracy (balanced accuracy) | Minimum of accuracy (balanced accuracy) | Maximum of accuracy (balanced accuracy) | |
---|---|---|---|---|
SVM (as in COMBI) | 0.59 (0.54) | 0.05 (0.06) | 0.41 (0.35) | 0.76 (0.71) |
DNN (as in DeepCOMBI) | 0.74 (0.74) | 0.07 (0.07) | 0.55 (0.50) | 0.97 (0.98) |