Timeline for Poisson binomial conjecture
Current License: CC BY-SA 4.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 23 at 1:11 | comment | added | Thomas Steinke | I'm glad it's useful! The next step would be to try defining a randomized function that acts on pairs of bits and evens out the probabilities to make it uniform. I didn't try working that out yet because this first step is already lossy. | |
| Aug 22 at 23:51 | comment | added | Aryeh Kontorovich | ok, I think I can make use of your ideas to get what I need (without proving the full conjecture). Would love to acknowledge you by name, if you'd be willing to let it be known. | |
| Aug 22 at 22:28 | comment | added | Aryeh Kontorovich | Very nice ideas there! I'd be happy with $(*)$ up to some multiplicative constant, such as 2 or even greater. Unfortunately, I do need the uniformity of $\bar p,\bar q$ -- which the $\hat p, \hat q$ doesn't quite give... But the idea to use the data processing inequality could be quite useful. I'll see if I can push it further... (note, however, that $(*)$ is definitely false for odd $n$ with constant 1) | |
| S Aug 22 at 20:56 | review | First answers | |||
| Aug 22 at 21:14 | |||||
| S Aug 22 at 20:56 | history | answered | Thomas Steinke | CC BY-SA 4.0 |