Table 9.
Confusion matrix and metrics of the classification with six classes, applied to the reference sample of 4XMM-DR10.
Truth → | AGN | Star | V/Y* | CV | gXB | eXB |
---|---|---|---|---|---|---|
→ AGN | 18234 | 9 | 2 | 147 | 11 | 18 |
→ Star | 12 | 4605 | 137 | 5 | 2 | 2 |
→ V/Y* | 1 | 45 | 53 | 3 | 0 | 0 |
→ CV | 1 | 0 | 1 | 85 | 6 | 0 |
→ gXB | 2 | 0 | 1 | 2 | 55 | 0 |
→ eXB | 61 | 0 | 0 | 8 | 0 | 429 |
Total | 18311 | 4659 | 194 | 250 | 75 | 449 |
Total cl. | 18421 | 4659 | 101 | 60 | 487 | 99 |
recall | 99.6 | 98.8 | 27.3 | 34.0 | 73.3 | 95.5 |
precision | 97.6 | 93.6 | 68.6 | 87.9 | 95.6 | 95.3 |
f1 score | .986 | .961 | .391 | .490 | .830 | .954 |
Notes. The first six rows show the number counts of objects classified as AGN, normal stars, variable or young stars, CVs, galactic XRBs, and extragalactic XRBs, respectively. The last three rows give the recall rate, the true positive rate (precision), and the f1 score associated with each class.
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.