Nonconvex regularized classification on Handwritten Digits 8x8
1.07Final Gradient Norm SquaredLT-ADMM
Evaluation results
| Method | Links | |
|---|---|---|
| LT-ADMMBatch size=1, Graph=Ring (N=10), Local training steps (τ)=2, Variance Reduction=No2025.01 | 1.07 | |
| LEDBatch size=1, Graph=Ring (N=10), Local training steps (τ)=2, Variance Reduction=No2025.01 | 1.29 | |
| K-GTBatch size=1, Graph=Ring (N=10), Local training steps (τ)=2, Variance Reduction=No2025.01 | 2.01 |