Timeline for (How) should multiple comparison correction be applied for hypothesis testing of cross validation performance?
Current License: CC BY-SA 4.0
21 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
Mar 30, 2023 at 11:23 | vote | accept | M.G.Poirot | ||
Mar 22, 2023 at 1:16 | history | edited | M.G.Poirot | CC BY-SA 4.0 |
added 2 characters in body
|
Mar 21, 2023 at 18:49 | history | edited | M.G.Poirot | CC BY-SA 4.0 |
deleted 4 characters in body
|
Mar 21, 2023 at 18:35 | history | edited | M.G.Poirot | CC BY-SA 4.0 |
added 34 characters in body
|
Mar 21, 2023 at 17:27 | history | edited | M.G.Poirot | CC BY-SA 4.0 |
deleted 418 characters in body
|
Mar 21, 2023 at 17:22 | history | edited | M.G.Poirot | CC BY-SA 4.0 |
deleted 418 characters in body
|
Mar 21, 2023 at 17:00 | history | edited | Sycorax♦ | CC BY-SA 4.0 |
appended answer 610189 as supplemental
|
S Mar 21, 2023 at 15:11 | history | bounty ended | M.G.Poirot | ||
S Mar 21, 2023 at 15:11 | history | notice removed | M.G.Poirot | ||
Mar 20, 2023 at 17:00 | history | edited | Firebug | CC BY-SA 4.0 |
edited title
|
Mar 20, 2023 at 10:35 | answer | added | Firebug | timeline score: 1 | |
Mar 20, 2023 at 10:23 | comment | added | Firebug | See this answer stats.stackexchange.com/a/536162/60613 | |
Mar 20, 2023 at 0:11 | comment | added | M.G.Poirot | Not at all: it does neither touch upon statistical testing nor MCC. In addition, as opposed to this question, the crux of that question is on the "nested" part of a CV-scheme, which is not a part of this question. | |
Mar 18, 2023 at 19:06 | comment | added | David B | Ok, then it seems like your question is a duplicate of stats.stackexchange.com/questions/250282/…. Is your question answered there? | |
Mar 18, 2023 at 16:58 | comment | added | M.G.Poirot |
It is not nested, it is simply 5-fold, 5-repeat CV, literally the most simple output of sklearn.model_selection.RepeatedKFold : Thus, performance_model_A is an 1D array of length n_splits × n_folds .
|
|
Mar 18, 2023 at 13:17 | comment | added | David B | Can you give more detail on your CV procedure? Is it nested CV? For example, with 5 splits, is it 5-fold CV, or do you hold-out 1/5th of the data, train (with some unspecified amount of CV) in the remaining 4/5th, and then test in the held-out 1/5th? | |
S Mar 18, 2023 at 10:57 | history | bounty started | M.G.Poirot | ||
S Mar 18, 2023 at 10:57 | history | notice added | M.G.Poirot | Authoritative reference needed | |
Mar 18, 2023 at 10:54 | history | edited | M.G.Poirot | CC BY-SA 4.0 |
added 21 characters in body
|
Mar 16, 2023 at 16:53 | answer | added | cbeleites | timeline score: 4 | |
Mar 15, 2023 at 20:26 | history | asked | M.G.Poirot | CC BY-SA 4.0 |