Table 1.
Aggregated results for all the metrics and all the synthetic clusters, for the six pyUPMASK clustering methods used, as percentage of results where pyUPMASK outperformed UPMASK, and vice versa, respectively.
| Method | pyUPMASK | UPMASK |
|---|---|---|
| min | max | min | max | |
| VOR | 66 | 29 |
| 55 (BSL) | 78 (TPR9) | 13 (TPR9) | 42 (PPV5) | |
| KNN | 68 | 27 |
| 57 (BSL) | 85 (TPR9) | 8 (TPR9) | 40 (PPV5) | |
| AGG | 72 | 24 |
| 59 (BSL) | 90 (TPR9) | 4 (TPR9) | 38 (PPV5) | |
| MBK | 77 | 16 |
| 51 (TPR9) | 90 (MCC5) | 8 (HMS) | 36 (TPR9) | |
| KMS | 79 | 15 |
| 66 (TPR9) | 88 (HMS) | 8 (PPV9) | 22 (TPR9) | |
| GMM | 83 | 11 |
| 74 (TPR9) | 93 (HMS) | 5 (HMS) | 16 (LSR) |
Notes. The missing percentage to add up to 100 corresponds to ties. The second rows for each method show the minimum and maximum percentage values obtained for any single metric (shown in parenthesis), for that method.
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.