Timeline for Does increasing number of observations lead to the decreasing of Mean Square Error of consistent estimators?
Current License: CC BY-SA 4.0
3 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 19, 2023 at 22:19 | comment | added | Alecos Papadopoulos | @whn The whole point is that to have a divergent $E(\Delta_n \mid \Delta_n \geq \epsilon)$, and in fact "so" divergent that it overcomes the plunge of $\Pr(\Delta_n \geq \epsilon)$ towards zero, is something that does not happen except in artificially constructed examples (caveat: all discussion here is subject to all these expected values existing of course). Think of what kind of distributions may $\Delta_n$ follow. | |
| Oct 18, 2023 at 19:04 | comment | added | whn | Thank you very much for your very delailed response ! However I don't understand why MSE should have certainly a decreasing value. I mean or $E(\Delta_n | \Delta_n > \epsilon )$ is bounded and then the estimator is MSE consistent or $E(\Delta_n | \Delta_n > \epsilon )$ diverge and then we cannot conclude anything on the decreasing of the MSE. I try to find something with successive decreasing values of epsilon but I don't succeed yet. | |
| Oct 13, 2023 at 2:55 | history | answered | Alecos Papadopoulos | CC BY-SA 4.0 |