Behavioral prediction on CCNP Out-of-domain
0.33Externalizing Pearson CorrelationNERVE
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| NERVEProtocol=SSL pre-training + downstream KRR, Input=FC matrices2026.05 | 0.33 | 0.08 | 0.13 | |
| GraphMAEProtocol=SSL pre-training + downstream KRR, Input=FC matrices2026.05 | 0.22 | 0.07 | 0.12 | |
| BrainGSLsProtocol=SSL pre-training + downstream KRR, Input=FC matrices2026.05 | 0.22 | — | — | |
| BrainNetTFProtocol=Supervised end-to-end, Input=BOLD time series2026.05 | 0.21 | 0.01 | 0.05 | |
| BrainGNNProtocol=Supervised end-to-end, Input=FC matrices2026.05 | 0.16 | 0.17 | 0.04 | |
| GATEProtocol=SSL pre-training + downstream KRR, Input=BOLD time series2026.05 | 0.14 | 0.04 | 0.11 | |
| BrainMassProtocol=SSL pre-training + downstream KRR, Input=FC matrices2026.05 | 0.14 | 0.04 | 0.13 | |
| BrainNetCNNProtocol=Supervised end-to-end, Input=FC matrices2026.05 | 0.1 | 0.15 | 0.01 |