Logistic Regression on Identity Σa
96.5Coverage (%)Weighted sample covariance estimator
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Weighted sample covariance estimatord (Dimension)=100, Optimization Method=Sketched Newton Method, tau (τ)=10, Estimator (Ξt vs Ξ¯t)=Ξt2025.02 | 96.5 | 0.195 | |
| Weighted sample covariance estimatord (Dimension)=20, Optimization Method=Sketched Newton Method, tau (τ)=∞, Estimator (Ξt vs Ξ¯t)=Ξt2025.02 | 95.5 | 0.038 | |
| Weighted sample covariance estimatord (Dimension)=40, Optimization Method=Sketched Newton Method, tau (τ)=20, Estimator (Ξt vs Ξ¯t)=Ξt2025.02 | 95.5 | 0.073 | |
| Sketched Newton Methodd (Dimension)=20, Optimization Method=Sketched Newton Method, tau (τ)=∞, Estimator (Ξt vs Ξ¯t)=Ξ¯t2025.02 | 95 | 0.04 | |
| Weighted sample covariance estimatord (Dimension)=20, Optimization Method=Sketched Newton Method, tau (τ)=10, Estimator (Ξt vs Ξ¯t)=Ξt2025.02 | 94.5 | 0.021 | |
| Weighted sample covariance estimatord (Dimension)=40, Optimization Method=Sketched Newton Method, tau (τ)=10, Estimator (Ξt vs Ξ¯t)=Ξt2025.02 | 94 | 0.052 | |
| SGDd (Dimension)=20, Optimization Method=SGD, tau (τ)=N/A, Estimator (Ξt vs Ξ¯t)=Ξ¯t2025.02 | 88 | -0.262 |