ProbXiv
Explorersign in
Problem archiveProblem record

Statement

Let G=FpG={\mathbf{F}}_p or G=RG=\mathbf{R}, let 0<ϵ<1/20<\epsilon<1/2, and let XX and YY be independent discrete random variables taking values in GG, under the hypotheses of Theorem 1. The theorem currently gives

H⁡(X+Y)≥12H⁡(X)+12H⁡(Y)+18−ϵ.\operatorname{\mathbf{H}}(X+Y)\geq \frac{1}{2}\operatorname{\mathbf{H}}(X)+\frac{1}{2}\operatorname{\mathbf{H}}(Y)+\frac{1}{8}-\epsilon.

Entropy sum lower-bound conjecture. The constant 1/81/8 can be improved to 1/21/2, meaning that

H⁡(X+Y)≥12H⁡(X)+12H⁡(Y)+12−ϵ.\operatorname{\mathbf{H}}(X+Y)\geq \frac{1}{2}\operatorname{\mathbf{H}}(X)+\frac{1}{2}\operatorname{\mathbf{H}}(Y)+\frac{1}{2}-\epsilon.

The value 1/81/8 is known for the stated non-identically distributed setting, whereas the value 1/21/2 holds when the variables are identically distributed and is suggested by Shannon's continuous entropy power inequality. Improving the constant in the general independent setting remains open.

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.