Black-box Optimization on BBOB surrogate 10-dimensional (out-of-distribution)
71.65Rastrigin Function ValueGradient-derived Evolutionary Composite (GEC)
Evaluation results
| Method | Links | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Gradient-derived Evolutionary Composite (GEC)2025.12 | 71.65 | 91.91 | — | 93.69 | 10.87 | 1,095,000 | 1.337 | 69.93 | 7.496 | 2.595 | 10.44 | 0.3979 | 14.58 | 15.36 | 1.782 | 64.57 | 13.48 | |
| B2OPT2025.12 | 88.77 | 141.6 | — | 518.3 | 33.59 | 9,190,000 | 3.331 | 100.8 | 13.67 | 3.199 | 8.515 | 0.7589 | 37.27 | 27.59 | 2.072 | 80.35 | 18.58 | |
| DEDQN2025.12 | 178.7 | 432.5 | — | 6,970 | 11,150 | 26,430,000 | 10.09 | 176.2 | 23 | 9.034 | 35.02 | 10.58 | 47.18 | 47.9 | 3.236 | 140.3 | 67.88 | |
| LGA2025.12 | 179.4 | 398.3 | — | 23,670 | 18,470 | 32,950,000 | 12.54 | 161.4 | 16.53 | 9.011 | 32.39 | 13.07 | 44.06 | 62.17 | 1.991 | 196.7 | 107.4 | |
| GLHF2025.12 | 244.6 | 864.9 | — | 19,120 | 11,270 | 37,110,000 | 18.06 | 221.5 | 16.83 | 8.558 | 36.09 | 8.791 | 44.41 | 52.12 | 2.869 | 159.4 | 171.1 | |
| RNNOPT2025.12 | 330.4 | 9,794 | — | 17,700 | 60.23 | 81,090,000 | 35.39 | 312.2 | 45.81 | 14.29 | 124.2 | 4.951 | 70.88 | 73.25 | 4.521 | 83.76 | 363 | |
| LES2025.12 | 2,605 | 2,659 | — | 3,211 | 1,538 | 15,540,000 | 1,906 | 2,590 | 1,908 | 103 | 1,230 | 2,000 | 934.6 | 1,349 | 1,002 | 568.8 | 527.6 |