Saved in:
Bibliographic Details
Main Authors: Lautrup, Anton Danholt, Rajabinasab, Muhammad, Hyrup, Tobias, Zimek, Arthur, Schneider-Kamp, Peter
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2507.19700
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916864940048384
author Lautrup, Anton Danholt
Rajabinasab, Muhammad
Hyrup, Tobias
Zimek, Arthur
Schneider-Kamp, Peter
author_facet Lautrup, Anton Danholt
Rajabinasab, Muhammad
Hyrup, Tobias
Zimek, Arthur
Schneider-Kamp, Peter
contents We propose a new framework for generating cross-sectional synthetic datasets via disjoint generative models. In this paradigm, a dataset is partitioned into disjoint subsets that are supplied to separate instances of generative models. The results are then combined post hoc by a joining operation that works in the absence of common variables/identifiers. The success of the framework is demonstrated through several case studies and examples on tabular data that helps illuminate some of the design choices that one may make. The principal benefit of disjoint generative models is significantly increased privacy at only a low utility cost. Additional findings include increased effectiveness and feasibility for certain model types and the possibility for mixed-model synthesis.
format Preprint
id arxiv_https___arxiv_org_abs_2507_19700
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Disjoint Generative Models
Lautrup, Anton Danholt
Rajabinasab, Muhammad
Hyrup, Tobias
Zimek, Arthur
Schneider-Kamp, Peter
Machine Learning
We propose a new framework for generating cross-sectional synthetic datasets via disjoint generative models. In this paradigm, a dataset is partitioned into disjoint subsets that are supplied to separate instances of generative models. The results are then combined post hoc by a joining operation that works in the absence of common variables/identifiers. The success of the framework is demonstrated through several case studies and examples on tabular data that helps illuminate some of the design choices that one may make. The principal benefit of disjoint generative models is significantly increased privacy at only a low utility cost. Additional findings include increased effectiveness and feasibility for certain model types and the possibility for mixed-model synthesis.
title Disjoint Generative Models
topic Machine Learning
url https://arxiv.org/abs/2507.19700