AD progression modeling on ADNI3 out-of-sample 2016-2022 (test)
27.1mF1Bayesian meta-learner
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Bayesian meta-learnerLast-Layer Inference (LLI)=false2026.06 | 27.1 | 0.74 | 0.44 | 0.22 | |
| Bayesian meta-learnerLast-Layer Inference (LLI)=true2026.06 | 27.1 | 0.74 | 0.44 | 0.18 | |
| Single-task learnerLast-Layer Inference (LLI)=false2026.06 | 6.03 | 1.51 | 1 | 0.52 | |
| Single-task learnerLast-Layer Inference (LLI)=true2026.06 | 5.89 | 1.18 | 0.78 | 0.38 |