Engagement Detection on Korea 2022 (test)
0.899F1 (weighted)non-FL bagging bootstrap
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| non-FL bagging bootstrapLearning Setting=Non-Federated, Model=Bagging Bootstrap2026.02 | 0.899 | 0.895 | 0.743 | |
| FedAwSLearning Setting=Federated, Model=bi-LSTM2026.02 | 0.786 | 0.718 | 0.709 | |
| FedAdamLearning Setting=Federated, Model=bi-LSTM2026.02 | 0.771 | 0.705 | 0.676 | |
| FedAvgLearning Setting=Federated, Model=bi-LSTM2026.02 | 0.764 | 0.686 | 0.702 | |
| non-FL bi-LSTMLearning Setting=Non-Federated, Model=bi-LSTM2026.02 | 0.751 | 0.671 | 0.753 | |
| FedProxLearning Setting=Federated, Model=bi-LSTM2026.02 | 0.718 | 0.628 | 0.737 | |
| MOONLearning Setting=Federated, Model=bi-LSTM2026.02 | 0.677 | 0.598 | 0.662 | |
| TurboSVMLearning Setting=Federated, Model=bi-LSTM2026.02 | 0.639 | 0.539 | 0.636 |