Table 1 Prediction performance at the optimal cutoff point for each input data.
From: Pilot study of eruption forecasting with muography using convolutional neural network
Method | Input region | Accuracy | Sensitivity | Specificity |
|---|---|---|---|---|
CNN model | Showa crater | 0.714 | 0.571 | 0.857 |
Minamidake craters | 0.667 | 0.523 | 0.810 | |
Surface | 0.524 | 0.333 | 0.714 | |
All segments | 0.643 | 0.619 | 0.667 | |
Rule-based method | 0.571 | 0.523 | 0.619 | |
SVM model (Showa crater region) | 0.619 | 0.667 | 0.571 | |
NN model (Showa crater region) | 0.548 | 0.286 | 0.810 | |
