Table 3.
Performance metrics of the SVM to classify linear regression values as belonging to nonflares (quiet active regions), flares happening in the next 24 hours, or flares happening within the next 24 hours to 6 days.
| Metric | Nonflaring | Flaring in 24 h | Flaring after 24 h | Combined |
|---|---|---|---|---|
| Accuracy | 0.93 | 0.88 | 0.83 | 0.87 |
| Precision | 0.93 | 0.57 | 0.66 | 0.81 |
| Recall | 0.94 | 0.62 | 0.60 | 0.81 |
| F1 | 0.94 | 0.59 | 0.62 | 0.81 |
| TSS | 0.84 | 0.54 | 0.50 | 0.71 |
| Primary Misclassification | Flaring after 24 h | Flaring after 24 h | Flaring in 24 h | – |
Notes. The “Combined” scores are calculated using micro-averaging. We further provide the primary class that was predicted in case of misclassifications for all classes.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.