Timeline for Maximum likelihood method vs. least squares method
Current License: CC BY-SA 3.0
16 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 29, 2023 at 23:20 | answer | added | ThomasHobbes | timeline score: 3 | |
| Dec 22, 2022 at 12:00 | history | tweeted | twitter.com/StackStats/status/1605895952576962560 | ||
| May 18, 2021 at 15:50 | answer | added | Lerner Zhang | timeline score: 0 | |
| Mar 5, 2017 at 2:02 | comment | added | Michael R. Chernick | @nail You are wrong to say that ML is a subset of LSE. If you look at the Gauss-Markov Theorem you will see that the estimators differ based on the distribution of the data. | |
| Mar 4, 2017 at 18:27 | answer | added | Lerner Zhang | timeline score: 44 | |
| Dec 12, 2015 at 21:40 | comment | added | whuber♦ | An answer is also provided at stats.stackexchange.com/questions/12562/…. | |
| Sep 13, 2015 at 11:00 | comment | added | nali | @TrynnaDoStat: No it's not the case. ML and LSE are not the same. In fact ML is an different approach which can emulate a subset of LSE and other estimators. See my answer for further details. | |
| Sep 13, 2015 at 1:24 | answer | added | nali | timeline score: 16 | |
| Mar 27, 2015 at 19:17 | comment | added | evros | thanks for all the replies. Now this makes sense. While searching for this topic on the net, I came across this article. Maybe this also helps:radfordneal.wordpress.com/2008/08/09/… | |
| Mar 27, 2015 at 17:41 | comment | added | whuber♦ | Search our site for the Gauss-Markov theorem. | |
| S Mar 27, 2015 at 15:25 | history | suggested | Richard Hardy | CC BY-SA 3.0 |
Added MLE and OLS tags, fixed grammar
|
| Mar 27, 2015 at 15:14 | review | Suggested edits | |||
| S Mar 27, 2015 at 15:25 | |||||
| Mar 27, 2015 at 15:07 | comment | added | TrynnaDoStat | Under normal error assumption, as is typically assumed in linear regression, the MLE and the LSE are the same! | |
| Mar 27, 2015 at 14:58 | review | First posts | |||
| Mar 27, 2015 at 15:12 | |||||
| Mar 27, 2015 at 14:57 | comment | added | Richard Hardy | You can use MLE in linear regression if you like. This can even make sense if the error distribution is non-normal and your goal is to obtain the "most likely" estimate rather than one which minimizes the sum of squares. | |
| Mar 27, 2015 at 14:54 | history | asked | evros | CC BY-SA 3.0 |