Table 2 Performance of algorithms over multiple runs using the 5-gene synthetic dataset, under three criteria: robustness over multiple runs, qualitative interactions and number of function calls performed
From: Comparison of evolutionary algorithms in gene regulatory network model inference
Criteria | PEACE1 | GA+ ES | GA+ ANN | GLSDC | DE+ AIC |
|---|---|---|---|---|---|
Robustness (Kinetic orders/Rate constants variance) | 0.25175/4.22818 | 0.4861/3.0170 | 0.07236 | 0.08449/2.0419 | 0.21534/6.41834 |
Identified interactions (Sensitivity/Specificity) | 0.55384/0.82702 | 0.6483/0.8902 | 0.74074/0.8125 | 0.72307/0.67837 | 0.58461/0.81081 |
Function calls | 1650000 | 3750000 | 2500 × 20000 ANN epochs | 100000 | 275000 |