Table 7 Precision of predictions on H1N1, H3N2 and H5N1 datasets for \(T \in \{5,10,15\}\).
From: Forecasting influenza hemagglutinin mutations through the lens of anomaly detection
Dataset | Method | TÂ =Â 5 | TÂ =Â 10 | TÂ =Â 15 | Mean |
|---|---|---|---|---|---|
H1N1 | Tempel41 | 0.8423 | 0.8445 | 0.8389 | 0.8419 |
Ours | \(0.8330 \pm 0.004\) | \(0.8292 \pm 0.004\) | \(0.8333 \pm 0.006\) | 0.8318 | |
H3N2 | Tempel41 | 0.7147 | 0.7300 | 0.7358 | 0.7268 |
Ours | \(0.5845 \pm 0.0004\) | \(0.5694 \pm 0.01\) | \(0.5786 \pm 0.01\) | 0.5775 | |
H5N1 | Tempel41 | 0.8721 | 0.8598 | 0.8791 | 0.7803 |
Ours | \(0.7530 \pm 0.02\) | \(0.7606 \pm 0.02\) | \(0.7176 \pm 0.02\) | 0.7437 |
