Table 3 Performance summary of 30 independent realizations for different regularization parameter \(\left( {{\varvec{\uplambda}}} \right)\) across all scenarios, showing success rates and the median of the objective function.
Scenario | \({{\varvec{\uplambda}}}\) | Realization success | Success rate (%) | Median ObjF (mgal) |
|---|---|---|---|---|
Single anomaly noise-free | 1e-5 | 25 | 83.3 | 1.38 × 10− 5 |
1e-4 | 28 | 93.3 | 1.71 × 10− 5 | |
1e-3 | 29 | 96.7 | 9.75 × 10− 6 | |
1e-2 | 28 | 93.3 | 7.35 × 10− 6 | |
1e-1 | 30 | 100 | 2.23 × 10− 5 | |
1 | 30 | 100 | 5.09 × 10− 4 | |
10 | 30 | 100 | 8.30 × 10− 3 | |
Single anomaly noise-contaminated | 1e-5 | 4 | 13.3 | 0.40 |
1e-4 | 7 | 23.3 | 0.40 | |
1e-3 | 24 | 80 | 0.40 | |
1e-2 | 29 | 96.7 | 0.40 | |
1e-1 | 30 | 100 | 0.40 | |
1 | 30 | 100 | 0.40 | |
10 | 30 | 100 | 0.40 | |
Multi anomaly noise-free | 1e-5 | 30 | 100 | 0.0551 |
1e-4 | 30 | 100 | 0.0587 | |
1e-3 | 30 | 100 | 0.0568 | |
1e-2 | 30 | 100 | 0.1191 | |
1e-1 | 30 | 100 | 0.2738 | |
1 | 30 | 100 | 0.3500 | |
10 | 30 | 100 | 0.3710 | |
Multi anomaly noise-contaminated | 1e-5 | 30 | 100 | 0.3717 |
1e-4 | 30 | 100 | 0.3933 | |
1e-3 | 30 | 100 | 0.3692 | |
1e-2 | 30 | 100 | 0.3751 | |
1e-1 | 30 | 100 | 0.3820 | |
1 | 30 | 100 | 0.3880 | |
10 | 30 | 100 | 0.4053 |
