Balabdaoui–Wellner conjecture on uniform Hermite spline interpolation error

Let kk be the spline parameter, let ff be a (k1)(k-1)-times differentiable function whose (k1)(k-1)st derivative has finite total variation, and let Ek(f)\mathcal{E}_k(f) denote the interpolation error in the Hermite interpolation problem via splines of degree 2k12k-1, with knots 0=τ0<τ1<<τ2k4<τ2k3=10=\tau_0<\tau_1<\cdots<\tau_{2k-4}<\tau_{2k-3}=1. Balabdaoui–Wellner conjecture. The interpolation error is bounded in the supremum norm independently of the locations of the knots; specifically, for u(0,1)u\in(0,1) and fu(t)=(tu)+k1/(k1)!f_u(t)=(t-u)^{k-1}_{+}/(k-1)!, there exists ck>0c_k>0 such that

supu(0,1)sup0<τ1<<τ2k4<1Ek(fu)ck.\sup_{u\in(0,1)}\sup_{0<\tau_1<\cdots<\tau_{2k-4}<1}\|\mathcal{E}_k(f_u)\|_\infty\leq c_k.

Such a bound is needed to establish that the distances between successive knots have stochastic order n1/(2k+1)n^{-1/(2k+1)} in the associated spline problem. The source does not provide evidence that the conjecture has been resolved.

Sources & referencesView supporting material

Primary source

Fadoua Balabdaoui and Jon A. Wellner, “Conjecture of error boundedness in a new Hermite interpolation problem via splines of odd-degree”, arXiv:math/0509160 (2005).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.