Gespeichert in:
Bibliographische Detailangaben
1. Verfasser: Wynne, George
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2507.18372
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916861643325440
author Wynne, George
author_facet Wynne, George
contents Publicly releasing the specification of a model with its trained parameters means an adversary can attempt to reconstruct information about the training data via training data reconstruction attacks, a major vulnerability of modern machine learning methods. This paper makes three primary contributions: establishing a mathematical framework to express the problem, characterising the features of the training data that are vulnerable via a maximum mean discrepancy equivalance and outlining a score matching framework for reconstructing data in both Bayesian and non-Bayesian models, the former is a first in the literature.
format Preprint
id arxiv_https___arxiv_org_abs_2507_18372
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle On Reconstructing Training Data From Bayesian Posteriors and Trained Models
Wynne, George
Machine Learning
Statistics Theory
Publicly releasing the specification of a model with its trained parameters means an adversary can attempt to reconstruct information about the training data via training data reconstruction attacks, a major vulnerability of modern machine learning methods. This paper makes three primary contributions: establishing a mathematical framework to express the problem, characterising the features of the training data that are vulnerable via a maximum mean discrepancy equivalance and outlining a score matching framework for reconstructing data in both Bayesian and non-Bayesian models, the former is a first in the literature.
title On Reconstructing Training Data From Bayesian Posteriors and Trained Models
topic Machine Learning
Statistics Theory
url https://arxiv.org/abs/2507.18372