Variational Inference on DS1
0.15Evaluation Time (min)SBN
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SBNtraining_iterations=400000, particles_per_iteration=10, evaluation_method=Importance Sampling, evaluation_particles=10002023.10 | 0.15 | 659.3 | |
| SBN*early_stopping=surpass CSMC baseline, evaluation_method=Importance Sampling, evaluation_particles=10002023.10 | 0.15 | 10.2 | |
| ARTreetraining_iterations=400000, particles_per_iteration=10, evaluation_method=Importance Sampling, evaluation_particles=10002023.10 | 0.41 | 3,740.8 | |
| ARTree*early_stopping=surpass CSMC baseline, evaluation_method=Importance Sampling, evaluation_particles=10002023.10 | 0.41 | 79.5 | |
| VaiPhytraining_iterations=200, particles_per_iteration=128, evaluation_method=Importance Sampling, evaluation_particles=10002023.10 | 1.6 | 45.1 | |
| VCSMCtraining_iterations=100, particles_per_iteration=2048, evaluation_method=Sequential Monte Carlo, evaluation_particles=20482023.10 | 2.4 | 248.3 | |
| phi-CSMCtraining=None (direct ML estimation), evaluation_method=Sequential Monte Carlo, evaluation_particles=20482023.10 | 102.2 | — |