ProbXiv
Explorersign in
Problem archiveProblem record

Statement

Let pp satisfy 1/2≤p≤11/2\leq p\leq 1, and let q00,q01,q10,q11≥0q_{00},q_{01},q_{10},q_{11}\geq 0 satisfy

q00+q01+q10+q11=1.q_{00}+q_{01}+q_{10}+q_{11}=1.

Bucketing-code exponent inequality. The following inequality should hold:

2p[q00ln⁡2q00p+q01ln⁡2q011−p+q10ln⁡2q101−p+q11ln⁡2q11p]≥2p\left[q_{00}\ln\frac{2q_{00}}{p}+q_{01}\ln\frac{2q_{01}}{1-p}+q_{10}\ln\frac{2q_{10}}{1-p}+q_{11}\ln\frac{2q_{11}}{p}\right] \geq (q00+q01)ln⁡(2(q00+q01))+(q10+q11)ln⁡(2(q10+q11))(q_{00}+q_{01})\ln\bigl(2(q_{00}+q_{01})\bigr)+(q_{10}+q_{11})\ln\bigl(2(q_{10}+q_{11})\bigr) +(q00+q10)ln⁡(2(q00+q10))+(q10+q11)ln⁡(2(q10+q11)).{}+(q_{00}+q_{10})\ln\bigl(2(q_{00}+q_{10})\bigr)+(q_{10}+q_{11})\ln\bigl(2(q_{10}+q_{11})\bigr).

This inequality would establish that the exponent 1/p1/p in the stated bucketing-code construction is optimal for the symmetric nearest-neighbor example. The source presents it as the single inequality reducing optimality to the direct lower bound, but gives no resolution here.

Record

Comments

Nothing has been published against this problem yet, and nothing has been said about it.

Sign in with an institutional address to take part in the discussion. Reading every thread stays open to everyone.

Sign in

Solve with an agent

Open the statement in a chat, with the problem and the ground rules already written into the prompt.

This opens a third-party site. Nothing is posted back to ProbXiv and nothing you write there is recorded here — what a model gives you is an attempt, which a person still has to check.