Timeline for Ridge, lasso and elastic net
Current License: CC BY-SA 3.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 18, 2021 at 0:21 | comment | added | John Vandivier | total non-expert here, but I think the idea is that X1 and X2 may be significantly indifferent explainers of Y, but the tiny unimportant edge of one over the other leads lasso to select it a floor the other, where Ridge would preserve them both. As such, lasso results may be harder to reproduce or cross validate. This MIGHT be consistent with why we call lasso an unstable selector. (Would love someone else to confirm) | |
| Sep 2, 2018 at 8:25 | comment | added | meTchaikovsky | @ balaks for the second point that you made, what does it mean of 'one does not know which variable gets picked'? Did you mean LASSO is indifferent, so it kind of randomly picks one so we don't really know which one is the best? | |
| Oct 18, 2016 at 17:18 | history | edited | balaks | CC BY-SA 3.0 |
edited body
|
| Oct 14, 2016 at 13:45 | review | Late answers | |||
| Oct 14, 2016 at 13:57 | |||||
| Oct 14, 2016 at 13:27 | review | First posts | |||
| Oct 14, 2016 at 13:50 | |||||
| Oct 14, 2016 at 13:22 | history | answered | balaks | CC BY-SA 3.0 |