Timeline for Log-linear and GLM (Poisson) regression
Current License: CC BY-SA 4.0
12 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 16, 2021 at 6:00 | history | tweeted | twitter.com/StackStats/status/1471359462418092037 | ||
| Dec 13, 2021 at 8:01 | comment | added | Mat | Thanks @Roland, this is very helpful! | |
| Dec 13, 2021 at 7:54 | comment | added | Roland | If you expect larger uncertainty for higher amounts (which seems very reasonable in accounting), your approach is very suitable. Of course, you need to consider that you are assuming an exponential relationship on the original scale. | |
| Dec 13, 2021 at 6:36 | comment | added | Mat | Thank you @Glen_b! I just did it and actually the results are very similar to the ones I got with my first approach. In this situation, are there any reasons why I should favor more one model than the other? | |
| Dec 13, 2021 at 6:03 | history | became hot network question | |||
| Dec 12, 2021 at 22:41 | comment | added | Glen_b | Poisson regression is for counts. A more suitable GLM might perhaps be a Gamma with log-link. | |
| Dec 12, 2021 at 19:45 | vote | accept | Mat | ||
| Dec 12, 2021 at 19:20 | answer | added | philbo_baggins | timeline score: 4 | |
| Dec 12, 2021 at 18:57 | answer | added | EdM | timeline score: 7 | |
| S Dec 12, 2021 at 18:01 | history | suggested | Pitouille |
Adding regression tag
|
|
| Dec 12, 2021 at 18:01 | review | Suggested edits | |||
| S Dec 12, 2021 at 18:01 | |||||
| Dec 12, 2021 at 17:57 | history | asked | Mat | CC BY-SA 4.0 |