Timeline for Loss function for estimating the conditional variance by fitting $y_i^2$
Current License: CC BY-SA 4.0
4 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 8, 2023 at 18:40 | comment | added | Richard Hardy | @JoseOrtiz3, if you are not sure about NLS, just do maximum likelihood. (Gradient boosting is not an estimator, so it does not really compete with these two.) | |
| Jun 8, 2023 at 18:17 | comment | added | JoseOrtiz3 | Sorry for the late feedback. I think the model scaling a unit normal is interesting, but I'm unclear on what you mean by estimating the parameters by nonlinear least squares. I derived a correct result in the gradient boosting case but haven't made a post yet. | |
| May 5, 2023 at 10:38 | comment | added | Richard Hardy | I probably do not fully understand your question, but here is something. | |
| May 5, 2023 at 10:37 | history | answered | Richard Hardy | CC BY-SA 4.0 |