Constrained Optimization on Concentric Circles Linear objective
100Feasibility (%)Projected Gradient
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Projected GradientMethod Category=Classical2026.04 | 100 | 14.83 | 1.16 | |
| FABMethod Category=ML-Based2026.04 | 100 | 0.38 | 1.48 | |
| FSNetMethod Category=ML-Based2026.04 | 98.5 | 3.04 | 1.72 | |
| FAB: No DξMethod Category=ML-Based2026.04 | 98.5 | 0.72 | 1.23 | |
| FAB: No Hinge LossMethod Category=ML-Based2026.04 | 98.1 | 0.58 | 1.48 | |
| FAB: Phase 1 OnlyMethod Category=ML-Based2026.04 | 98 | 0.54 | 1.12 | |
| FAB: 3 Decoder NetsMethod Category=ML-Based2026.04 | 96.9 | 0.61 | 1.6 | |
| Penalty NNMethod Category=ML-Based2026.04 | 96.2 | 0.14 | 1.06 | |
| Homeomorphic ProjectionMethod Category=ML-Based2026.04 | 93.4 | 168 | 2.56 | |
| FAB: 2 Decoder NetsMethod Category=ML-Based2026.04 | 92.7 | 0.49 | 1.52 | |
| Interior PointMethod Category=Classical2026.04 | 89.9 | 39.76 | 1.78 | |
| Penalty MethodMethod Category=Classical2026.04 | 42.5 | 39.15 | 1.35 | |
| Augmented LagrangianMethod Category=Classical2026.04 | 42.1 | 52.09 | 1.39 |