Timeline for Biased Data in Machine Learning
Current License: CC BY-SA 3.0
20 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 10, 2017 at 21:45 | comment | added | seanv507 | This is called the exploration-exploitation dilemma. | |
| Sep 10, 2017 at 21:29 | answer | added | Dynamic Stardust | timeline score: 1 | |
| Sep 7, 2017 at 10:11 | vote | accept | Laksan Nathan | ||
| Sep 7, 2017 at 10:11 | vote | accept | Laksan Nathan | ||
| Sep 7, 2017 at 10:11 | |||||
| Sep 6, 2017 at 14:37 | comment | added | Laksan Nathan | Preparing my presentation right now. This analogy really comes in handy, thanks! | |
| Sep 6, 2017 at 13:07 | comment | added | MSalters | When I explain this problem to non-experts, I draw a cloud (the reality), and a polygon approximating the cloud (the model). I show the false positive errors and false negative errors. It's visually clear that I need both errors to improve the model, so to approximate the cloud better. | |
| Sep 6, 2017 at 8:47 | answer | added | oDDsKooL | timeline score: 2 | |
| Sep 6, 2017 at 7:50 | answer | added | RandomStats | timeline score: 2 | |
| Sep 6, 2017 at 4:09 | history | tweeted | twitter.com/StackStats/status/905281901950328832 | ||
| Sep 5, 2017 at 20:41 | answer | added | alegarra | timeline score: 1 | |
| Sep 5, 2017 at 18:53 | answer | added | Jim K. | timeline score: 2 | |
| Sep 5, 2017 at 17:13 | vote | accept | Laksan Nathan | ||
| Sep 7, 2017 at 10:11 | |||||
| Sep 5, 2017 at 17:13 | vote | accept | Laksan Nathan | ||
| Sep 5, 2017 at 17:13 | |||||
| Sep 5, 2017 at 14:52 | vote | accept | Laksan Nathan | ||
| Sep 5, 2017 at 14:53 | |||||
| Sep 5, 2017 at 14:51 | comment | added | Laksan Nathan | This is really good to know. Maybe I can even set up things to do the same. | |
| Sep 5, 2017 at 14:23 | answer | added | rinspy | timeline score: 13 | |
| Sep 5, 2017 at 14:16 | comment | added | Matthew Drury | The way this is generally handled in credit risk assessment is by not filtering a certain proportion of applicants by the rules. A small number of applicants are randomly admitted, and flagged as such. | |
| Sep 5, 2017 at 13:39 | answer | added | bibliolytic | timeline score: 4 | |
| Sep 5, 2017 at 13:15 | history | edited | Laksan Nathan | CC BY-SA 3.0 |
added 2 characters in body
|
| Sep 5, 2017 at 13:05 | history | asked | Laksan Nathan | CC BY-SA 3.0 |