Table 1
Supervised classification results in the latent space.
| Latent variables from: | F1−score [%] | |||
|---|---|---|---|---|
| Logistic regressor | k-NN classifier | |||
| Macro | Weighted | Macro | Weighted | |
| XP spectra | 39.9(1.1) | 55.1(0.9) | 52.3(0.8) | 68.1(0.5) |
| Δm|Δt LC | 28.4(0.5) | 39.7(0.8) | 27.8(0.7) | 51.6(0.1) |
| Folded LC | 36.1(0.1) | 47.9(0.1) | 35.8(1.2) | 55.2(0.1) |
| Δm|Δt&Fold | 43.1(0.7) | 66.6(0.5) | 39.3(0.4) | 64.5(0.3) |
| All inputs | 68.2(0.4) | 79.9(0.4) | 66.1(0.3) | 80.5(0.2) |
Notes. Average F1−scores on the test set for logistic regression and k-nearest neighbours classifiers, trained using latent variables from XP mean spectra, folded light curves (LC), Δm|Δt from LC and combinations between them. Standard deviations are shown in parentheses. A higher F1−score implies a better classification.
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.