English

Probabilistic divide-and-conquer: deterministic second half

Probability 2016-09-15 v2

Abstract

We present a probabilistic divide-and-conquer (PDC) method for \emph{exact} sampling of conditional distributions of the form L(XXE)\mathcal{L}( {\bf X}\, |\, {\bf X} \in E), where X{\bf X} is a random variable on X\mathcal{X}, a complete, separable metric space, and event EE with P(E)0\mathbb{P}(E) \geq 0 is assumed to have sufficient regularity such that the conditional distribution exists and is unique up to almost sure equivalence. The PDC approach is to define a decomposition of X\mathcal{X} via sets A\mathcal{A} and B\mathcal{B} such that X=A×B\mathcal{X} = \mathcal{A} \times \mathcal{B}, and sample from each separately. The deterministic second half approach is to select the sets A\mathcal{A} and B\mathcal{B} such that for each element aAa\in \mathcal{A}, there is only one element baBb_a \in \mathcal{B} for which (a,ba)E(a,b_a)\in E. We show how this simple approach provides non-trivial improvements to several conventional random sampling algorithms in combinatorics, and we demonstrate its versatility with applications to sampling from sufficiently regular conditional distributions.

Keywords

Cite

@article{arxiv.1411.6698,
  title  = {Probabilistic divide-and-conquer: deterministic second half},
  author = {Stephen DeSalvo},
  journal= {arXiv preprint arXiv:1411.6698},
  year   = {2016}
}

Comments

28 pages

R2 v1 2026-06-22T07:10:53.100Z