Timeline for What would I prefer - an over-fitted model or a less accurate model?
Current License: CC BY-SA 4.0
4 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 1, 2020 at 1:51 | history | edited | Ben Reiniger♦ | CC BY-SA 4.0 |
additional stats.SE question
|
| Feb 3, 2020 at 21:08 | history | edited | Ben Reiniger♦ | CC BY-SA 4.0 |
another reference link
|
| Jan 13, 2020 at 14:18 | comment | added | Peteris | There's an interesting paper arxiv.org/abs/1812.11118v2 with a claim that with a sufficiently powerful network you get 'beyond overfitting' as the test accuracy starts to improve (due to regularization and other prior structural bias) after overfitting while the train accuracy stays at 100%. | |
| Jan 13, 2020 at 2:54 | history | answered | Ben Reiniger♦ | CC BY-SA 4.0 |