Timeline for How to improve F1 score with skewed classes?
Current License: CC BY-SA 4.0
8 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 15, 2020 at 22:01 | comment | added | Dave | Why do you make point #3? | |
| Oct 14, 2020 at 9:02 | comment | added | Green Falcon | Do you usually fine tune threshold for F1 score after training a model is completed? | |
| Mar 18, 2020 at 13:55 | history | edited | Sandeep S. Sandhu | CC BY-SA 4.0 |
Caution on the use of cutoffs
|
| Apr 21, 2017 at 8:46 | vote | accept | Gian Segato | ||
| Apr 20, 2017 at 17:56 | history | edited | Sandeep S. Sandhu | CC BY-SA 3.0 |
Added description for choosing a good cutoff
|
| Apr 20, 2017 at 17:51 | history | edited | Sandeep S. Sandhu | CC BY-SA 3.0 |
Added description for choosing a good cutoff
|
| Apr 20, 2017 at 17:27 | comment | added | Matthew Drury | It is important that your learning algorithm predict class probabilities, and that a threshold is set for hard classification based on the risks of false negatives and false positives. I do not think your answer is complete without mentioning this. | |
| Apr 20, 2017 at 16:55 | history | answered | Sandeep S. Sandhu | CC BY-SA 3.0 |