Regression on Stroke Dataset (subject-wise cross-validation)
2.82RMSEES-VAE + k-NN
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ES-VAE + k-NNInput Representation=Tangent Vector2026.05 | 2.82 | 74 | 0.86 | |
| Tangent PCA + k-NNInput Representation=Tangent Vector2026.05 | 3.03 | 70 | 0.84 | |
| LSTMInput Representation=Tangent Vector2026.05 | 3.21 | 66 | 0.81 | |
| TransformerInput Representation=Tangent Vector2026.05 | 3.23 | 66 | 0.81 | |
| LSTMInput Representation=Raw Skeleton2026.05 | 3.31 | 64 | 0.81 | |
| Sparse ST-GCNInput Representation=Tangent Vector2026.05 | 3.36 | 63 | 0.8 | |
| TCNInput Representation=Tangent Vector2026.05 | 3.39 | 62 | 0.82 | |
| ST-GCNInput Representation=Tangent Vector2026.05 | 3.51 | 60 | 0.77 | |
| Hyper-GCNInput Representation=Raw Skeleton2026.05 | 3.57 | 58 | 0.77 | |
| VAE + k-NNInput Representation=Joint Angle2026.05 | 3.71 | 55 | 0.76 | |
| TCNInput Representation=Raw Skeleton2026.05 | 3.82 | 52 | 0.79 | |
| Sparse ST-GCNInput Representation=Raw Skeleton2026.05 | 3.83 | 52 | 0.74 | |
| Hyper-GCNInput Representation=Tangent Vector2026.05 | 3.86 | 51 | 0.74 | |
| ST-GCNInput Representation=Raw Skeleton2026.05 | 3.88 | 51 | 0.72 | |
| TransformerInput Representation=Raw Skeleton2026.05 | 3.93 | 50 | 0.76 | |
| PCA + k-NNInput Representation=Joint Angle2026.05 | 4.13 | 44 | 0.76 | |
| VAE + k-NNInput Representation=Raw Skeleton2026.05 | 4.33 | 39 | 0.62 | |
| PCA + k-NNInput Representation=Raw Skeleton2026.05 | 4.46 | 35 | 0.59 |