ProbXiv
Explorersign in
Problem archiveProblem record

Statement

Let β(n)=(β0(n),β1(n),…,βN(n))\beta^{(n)}=(\beta^{(n)}_0,\beta^{(n)}_1,\dots,\beta^{(n)}_N) be the coefficient vector and let β^(n)\hat{\beta}^{(n)} be its numerical approximation computed in Step 2. Let ε\varepsilon denote the machine accuracy, for example ε≈10−16\varepsilon\approx 10^{-16} in double precision, and let cc be the band limit.

Relative-accuracy conjecture. The coordinates of β(n)\beta^{(n)} satisfy, for every 1≤k≤N1\leq k\leq N,

∣βk(n)−β^k(n)βk(n)∣≤ε⋅log⁡(c).\left|\frac{\beta^{(n)}_k-\hat{\beta}^{(n)}_k}{\beta^{(n)}_k}\right|\leq\varepsilon\cdot\log(\sqrt{c}).

The conjecture asserts that even coordinates that are small in absolute value are computed with high relative accuracy. The source reports preliminary analysis and extensive numerical experiments supporting it, while stating that the proof remains a subject of ongoing research.

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.