Timeline for Total variation distance between multinomial laws
Current License: CC BY-SA 3.0
3 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 23, 2016 at 17:39 | comment | added | Yuval Peres | This is essentially correct, but there might be some confusion between the product measures $p^n$ and $q^n$, and the multinomial measures that are their projections. The reason the proof works is that a symmetry argument shows that the total variation distance is not changed by the projection. Working directly with the multinomial measures would require a proof of the Martingale property since the corresponding $\sigma$ fields are not increasing. | |
| Sep 20, 2016 at 13:52 | comment | added | Alainty | Ok, right! Thank you very much michael!! | |
| Sep 20, 2016 at 7:01 | history | answered | user83457 | CC BY-SA 3.0 |