Local differential precision operator conjecture for Gaussian Markov random fields on metric graphs

Let Γ\Gamma be a compact metric graph with edge set E\mathcal{E}. A Gaussian Markov random field of order pp is assumed to satisfy some mild regularity conditions. On each edge eeqinmathcalEe eqinmathcal{E}, consider a differential operator with coefficients kappai,ekappa_{i,e}, and let MM be a finite-rank operator describing the action at the vertices. Local precision operator conjecture. Any Gaussian Markov random field uu of order pp can be obtained as a Gaussian random field whose precision operator acts on each edge as

Q=i=02pκi,edidxei,Q = \sum_{i=0}^{2p} \kappa_{i,e} \frac{d^{i}}{dx_e^{i}},

where the coefficient functions κi,e\kappa_{i,e} are sufficiently nice, and whose action at the vertices is given by MM, with QQ self-adjoint and local and having compact resolvent. This proposes a general local-operator representation for Gaussian Markov random fields on compact metric graphs; it remains open under the stated mild regularity assumptions.

Sources & referencesView supporting material

Primary source

David Bolin, Alexandre B. Simas and Jonas Wallin, “Markov properties of Gaussian random fields on compact metric graphs”, arXiv:2304.03190 (2024).

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.