Bayesian linear regression on Bayesian linear regression 10D
36RMSE (ε=0.1)NA-DPVI (NUTS)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| NA-DPVI (NUTS)Approximation method=NUTS, delta=10^-5, Scaling factor=10^32024.10 | 36 | 35 | 27 | |
| NA-DPVI (Laplace)Approximation method=Laplace, delta=10^-5, Scaling factor=10^32024.10 | 78 | 98 | 60 | |
| Bernstein & Sheldondelta=10^-5, Scaling factor=10^32024.10 | 301 | 299 | 323 | |
| Naive (last iterate)Approximation method=last iterate, delta=10^-5, Scaling factor=10^32024.10 | 512 | 307 | 360 | |
| DPVImdelta=10^-5, Scaling factor=10^32024.10 | 584 | 584 | 584 |