Sparse Regression on gisette scale
0.52CPU TimeAPG+
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| APG+Sparsity level (s)=[1.5m], Tolerance (epsilon)=10^-62022.11 | 0.52 | 118 | 133 | 0.339 | |
| APG+Sparsity level (s)=m, Tolerance (epsilon)=10^-62022.11 | 0.58 | 78 | 247 | 0.499 | |
| APG-LL+Sparsity level (s)=m, Tolerance (epsilon)=10^-62022.11 | 0.99 | 330 | 144 | 0.341 | |
| APG-LL+Sparsity level (s)=2m, Tolerance (epsilon)=10^-62022.11 | 1.13 | 488 | 128 | 0.336 | |
| APG-LL+Sparsity level (s)=[1.5m], Tolerance (epsilon)=10^-62022.11 | 1.15 | 500 | 150 | 0.387 | |
| APG+Sparsity level (s)=[1.1m], Tolerance (epsilon)=10^-62022.11 | 1.2 | 200 | 203 | 0.347 | |
| APG+Sparsity level (s)=2m, Tolerance (epsilon)=10^-62022.11 | 1.32 | 357 | 143 | 0.319 | |
| APG-LL+Sparsity level (s)=[1.1m], Tolerance (epsilon)=10^-62022.11 | 1.59 | 449 | 203 | 0.508 | |
| APGSparsity level (s)=2m, Tolerance (epsilon)=10^-62022.11 | 5.03 | 1,443 | 0 | 0.282 | |
| PG-LLSparsity level (s)=2m, Tolerance (epsilon)=10^-62022.11 | 6.98 | 2,860 | 0 | 0.328 | |
| APGSparsity level (s)=[1.5m], Tolerance (epsilon)=10^-62022.11 | 7.4 | 2,057 | 0 | 0.284 | |
| PG-LLSparsity level (s)=[1.5m], Tolerance (epsilon)=10^-62022.11 | 7.88 | 3,242 | 0 | 0.347 | |
| APGSparsity level (s)=[1.1m], Tolerance (epsilon)=10^-62022.11 | 15.55 | 3,000 | 0 | 0.308 | |
| APGSparsity level (s)=m, Tolerance (epsilon)=10^-62022.11 | 16.5 | 3,326 | 0 | 0.298 | |
| PG-LLSparsity level (s)=[1.1m], Tolerance (epsilon)=10^-62022.11 | 18.91 | 4,462 | 0 | 0.359 | |
| PG-LLSparsity level (s)=m, Tolerance (epsilon)=10^-62022.11 | 21.09 | 5,364 | 0 | 0.367 | |
| PGSparsity level (s)=2m, Tolerance (epsilon)=10^-62022.11 | 23.08 | 10,000 | 0 | 0.237 | |
| PGSparsity level (s)=[1.5m], Tolerance (epsilon)=10^-62022.11 | 23.09 | 10,000 | 0 | 0.23 | |
| PGSparsity level (s)=m, Tolerance (epsilon)=10^-62022.11 | 35.83 | 10,000 | 0 | 0.226 | |
| PGSparsity level (s)=[1.1m], Tolerance (epsilon)=10^-62022.11 | 40.06 | 10,000 | 0 | 0.225 |