ProbXiv
Explorersign in
Problem archiveProblem record

Statement

Let X,Y∈{0,1}X,Y \in \{0,1\} have joint distribution

pX,Y(x,y)=[12(1−α)12α12α12(1−α)],p_{X,Y}(x,y) = \begin{bmatrix} \frac{1}{2}(1- \alpha) & \frac{1}{2}\alpha \\ \frac{1}{2}\alpha & \frac{1}{2}(1- \alpha) \\ \end{bmatrix},

then I(X;Y)=1−h(α)I(X;Y) = 1 - h(\alpha) for 0≤α≤1/20 \leq \alpha \leq 1/2 where h(α)≜−αlog⁡2α−(1−α)log⁡2(1−α).h(\alpha) \triangleq -\alpha \log_2 \alpha - (1 - \alpha) \log_2 (1 - \alpha). Note that I(X;Y)≜D(pX,Y∣∣pXpY)I(X;Y) \triangleq D(p_{X,Y} || p_X p_Y).

The conjecture, often referred to as the “Courtade-Kumar Conjecture” or the “Most Informative Function Conjecture”, is the following. Let (Xi,Yi)(X_i, Y_i) for 1≤i≤n1 \leq i \leq n be i.i.d. =d(X,Y)\overset{d}{=} (X,Y), then

I(f(X1n);Y1n)≤1−h(α)I(f(X_1^n); Y_1^n) \leq 1 - h(\alpha)

for any function f:{0,1}n→{0,1}.f:\{0,1\}^n \rightarrow \{0,1\}. We note that there is prior work on special cases.

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.