Tight simulation of a distribution using conditional samples

Fuente: arXiv
Saved in:
Bibliographic Details
Main Author: Adar, Tomer
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918453577777152
author Adar, Tomer
author_facet Adar, Tomer
contents We present an algorithm for simulating a distribution using prefix conditional samples (Adar, Fischer and Levi, 2024), as well as ``prefix-compatible'' conditional models such as the interval model (Cannone, Ron and Servedio, 2015) and the subcube model (CRS15, Bhattacharyya and Chakraborty, 2018). The sample complexity is $O(\log^2 N / \varepsilon^2)$ prefix conditional samples per query, which improves on the previously known $\tilde{O}(\log^3 N / \varepsilon^2)$ (Kumar, Meel and Pote, 2025). Moreover, our simulating distribution is $O(\varepsilon^2)$-close to the input distribution with respect to the Kullback-Leibler divergence, which is stricter than the usual guarantee of being $O(\varepsilon)$-close with respect to the total-variation distance. We show that our algorithm is tight with respect to the highly-related task of estimation: every algorithm that is able to estimate the mass of individual elements within $(1 \pm \varepsilon)$-multiplicative error must make $Ω(\log^2 N / \varepsilon^2)$ prefix conditional samples per element.
format Preprint
id arxiv_https___arxiv_org_abs_2506_18444
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Tight simulation of a distribution using conditional samples
Adar, Tomer
Data Structures and Algorithms
We present an algorithm for simulating a distribution using prefix conditional samples (Adar, Fischer and Levi, 2024), as well as ``prefix-compatible'' conditional models such as the interval model (Cannone, Ron and Servedio, 2015) and the subcube model (CRS15, Bhattacharyya and Chakraborty, 2018). The sample complexity is $O(\log^2 N / \varepsilon^2)$ prefix conditional samples per query, which improves on the previously known $\tilde{O}(\log^3 N / \varepsilon^2)$ (Kumar, Meel and Pote, 2025). Moreover, our simulating distribution is $O(\varepsilon^2)$-close to the input distribution with respect to the Kullback-Leibler divergence, which is stricter than the usual guarantee of being $O(\varepsilon)$-close with respect to the total-variation distance. We show that our algorithm is tight with respect to the highly-related task of estimation: every algorithm that is able to estimate the mass of individual elements within $(1 \pm \varepsilon)$-multiplicative error must make $Ω(\log^2 N / \varepsilon^2)$ prefix conditional samples per element.
title Tight simulation of a distribution using conditional samples
topic Data Structures and Algorithms
url https://arxiv.org/abs/2506.18444