Scaling Up Bayesian DAG Sampling

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Nikzad, Daniele, Zhilkin, Alexander, Harviainen, Juha, Kuipers, Jack, Moffa, Giusi, Koivisto, Mikko
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908618937335808
author Nikzad, Daniele
Zhilkin, Alexander
Harviainen, Juha
Kuipers, Jack
Moffa, Giusi
Koivisto, Mikko
author_facet Nikzad, Daniele
Zhilkin, Alexander
Harviainen, Juha
Kuipers, Jack
Moffa, Giusi
Koivisto, Mikko
contents Bayesian inference of Bayesian network structures is often performed by sampling directed acyclic graphs along an appropriately constructed Markov chain. We present two techniques to improve sampling. First, we give an efficient implementation of basic moves, which add, delete, or reverse a single arc. Second, we expedite summing over parent sets, an expensive task required for more sophisticated moves: we devise a preprocessing method to prune possible parent sets so as to approximately preserve the sums. Our empirical study shows that our techniques can yield substantial efficiency gains compared to previous methods.
format Preprint
id arxiv_https___arxiv_org_abs_2510_25254
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Scaling Up Bayesian DAG Sampling
Nikzad, Daniele
Zhilkin, Alexander
Harviainen, Juha
Kuipers, Jack
Moffa, Giusi
Koivisto, Mikko
Machine Learning
Artificial Intelligence
Bayesian inference of Bayesian network structures is often performed by sampling directed acyclic graphs along an appropriately constructed Markov chain. We present two techniques to improve sampling. First, we give an efficient implementation of basic moves, which add, delete, or reverse a single arc. Second, we expedite summing over parent sets, an expensive task required for more sophisticated moves: we devise a preprocessing method to prune possible parent sets so as to approximately preserve the sums. Our empirical study shows that our techniques can yield substantial efficiency gains compared to previous methods.
title Scaling Up Bayesian DAG Sampling
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2510.25254