Timeline for Standard ways to automatically remove incorrectly classified observations from a (mostly categorical) training dataset?
Current License: CC BY-SA 4.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Mar 30, 2022 at 9:24 | answer | added | Jon Nordby | timeline score: 0 | |
| Mar 30, 2022 at 9:12 | comment | added | Jon Nordby | Since this is dataset curation I would advise caution and go through manually as practically possible, to ensure that the labels are correct. And if you need to automate things, it should probably be automation of the things you would do to manually correct the labels - not a generic automagical fixer model | |
| Mar 30, 2022 at 9:10 | comment | added | Jon Nordby | What are the nature of your issues? Do you have labels y=A which should actually have been y=B? Or you want to drop out-of-distribution / data from classes that should not be included? | |
| Jan 20, 2022 at 12:28 | history | edited | Giuliano Mirabella | CC BY-SA 4.0 |
added 24 characters in body
|
| Jan 20, 2022 at 10:20 | history | asked | Giuliano Mirabella | CC BY-SA 4.0 |