Random expansions of trees with bounded height
Abstract
We consider a sequence of trees where, for some every has height at most and as the minimal number of children of a nonleaf tends to infinity. We can view every tree as a (first-order) -structure where is a signature with one binary relation symbol. For a fixed (arbitrary) finite and relational signature we consider the set of expansions of to and a probability distribution on which is determined by a (parametrized/lifted) Probabilistic Graphical Model (PGM) which can use the information given by . The kind of PGM that we consider uses formulas of a many-valued logic that we call with truth values in the unit interval . We also use to express queries, or events, on . With this setup we prove that, under some assumptions on , , and a (possibly quite complex) formula of , as , if are vertices of the tree then the value of will, with high probability, be almost the same as the value of , where is a ``simple'' formula the value of which can always be computed quickly (without reference to ), and itself can be found by using only the information that defines , and . A corollary of this, subject to the same conditions, is a probabilistic convergence law for -formulas.
Cite
@article{arxiv.2410.11775,
title = {Random expansions of trees with bounded height},
author = {Vera Koponen and Yasmin Tousinejad},
journal= {arXiv preprint arXiv:2410.11775},
year = {2025}
}
Comments
arXiv admin note: text overlap with arXiv:2401.04802 Author comment: This is a minor revision (but with some important corrections) of the first version