Table 10
Approximate time for feature extraction and training of algorithms.
| Method | Extraction of features | Training algorithm | Total time |
|---|---|---|---|
| This work CNN | 50 s | 13.9 min | 14.73 min |
| This work RF | 5.44 s | 1.48 s | 6.92 s |
| Aguirre RF | 11.5 days | 36 min | 11.52 days |
| Aguirre CNN | 30 min | 50 min | 1.33 h |
Notes. We analyze a sample of 90 928 stars using 24 cores for parallel feature extraction and an RTX 4060 GPU for training. For comparison, we reference the time calculated by Aguirre et al. (2019), who utilized 6 CPUs for parallel feature extraction and a GeForce GTX 1080 Ti GPU, For a original sample of 51 951 stars.
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.