Search for hot subdwarf stars from SDSS images using a deep learning method: SwinBayesNet

Open Access

Table 6

Performance comparison of using photometric magnitudes and images.

Model	Two-class classification		Three-class classification

	Accuracy	F1	Accuracy	F1
SVM	0.89	0.88	0.74	0.83
RF	0.89	0.88	0.79	0.84
XGBoost	0.90	0.89	0.78	0.82
LightGBM	0.90	0.89	0.78	0.84
CatBoost	0.90	0.88	0.78	0.85
Ours	0.92	0.90	0.82	0.89

Notes. Except for the ‘Ours’ model, which is built based on photometric images, all others are established using photometric magnitudes.

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.