Sparsity-constrained Optimization on gisette scale
0.07CPU TimeAPG+
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| APG+Sparsity level (s)=0.01m, Tolerance (epsilon)=10^-62022.11 | 0.07 | 11 | 18 | 0.464 | |
| APG-LL+Sparsity level (s)=0.01m, Tolerance (epsilon)=10^-62022.11 | 0.07 | 32 | 18 | 0.464 | |
| APG+Sparsity level (s)=0.05m, Tolerance (epsilon)=10^-62022.11 | 0.08 | 13 | 38 | 0.334 | |
| APG-LL+Sparsity level (s)=0.05m, Tolerance (epsilon)=10^-62022.11 | 0.08 | 37 | 23 | 0.337 | |
| APG+Sparsity level (s)=0.1m, Tolerance (epsilon)=10^-62022.11 | 0.08 | 13 | 41 | 0.26 | |
| APG-LL+Sparsity level (s)=0.1m, Tolerance (epsilon)=10^-62022.11 | 0.09 | 54 | 40 | 0.259 | |
| APG+Sparsity level (s)=0.5m, Tolerance (epsilon)=10^-62022.11 | 0.22 | 43 | 108 | 0.253 | |
| APG-LL+Sparsity level (s)=0.5m, Tolerance (epsilon)=10^-62022.11 | 0.48 | 271 | 149 | 0.258 | |
| PG-LLSparsity level (s)=0.01m, Tolerance (epsilon)=10^-62022.11 | 2.45 | 1,758 | 0 | 0.466 | |
| APGSparsity level (s)=0.05m, Tolerance (epsilon)=10^-62022.11 | 3.63 | 1,286 | 0 | 0.303 | |
| APGSparsity level (s)=0.01m, Tolerance (epsilon)=10^-62022.11 | 4.08 | 1,526 | 0 | 0.464 | |
| PG-LLSparsity level (s)=0.1m, Tolerance (epsilon)=10^-62022.11 | 5.37 | 3,294 | 0 | 0.238 | |
| APGSparsity level (s)=0.1m, Tolerance (epsilon)=10^-62022.11 | 7.83 | 2,697 | 0 | 0.212 | |
| PGSparsity level (s)=0.01m, Tolerance (epsilon)=10^-62022.11 | 13.35 | 10,000 | 0 | 0.465 | |
| PGSparsity level (s)=0.05m, Tolerance (epsilon)=10^-62022.11 | 14.7 | 10,000 | 0 | 0.304 | |
| APGSparsity level (s)=0.5m, Tolerance (epsilon)=10^-62022.11 | 14.97 | 3,922 | 0 | 0.292 | |
| PGSparsity level (s)=0.1m, Tolerance (epsilon)=10^-62022.11 | 15.63 | 10,000 | 0 | 0.243 | |
| PGSparsity level (s)=0.5m, Tolerance (epsilon)=10^-62022.11 | 23.63 | 10,000 | 0 | 0.22 | |
| PG-LLSparsity level (s)=0.05m, Tolerance (epsilon)=10^-62022.11 | 30.15 | 20,000 | 0 | 0.268 | |
| PG-LLSparsity level (s)=0.5m, Tolerance (epsilon)=10^-62022.11 | 50.85 | 20,000 | 0 | 0.364 |