menu
techminis

A naukri.com initiative

google-web-stories
Home

>

ML News

>

RelDiff: R...
source image

Arxiv

3d

read

326

img
dot

Image Credit: Arxiv

RelDiff: Relational Data Generative Modeling with Graph-Based Diffusion Models

  • RelDiff is a novel diffusion generative model designed for relational databases.
  • It explicitly models the foreign key graph structure of relational databases.
  • RelDiff combines a graph-conditioned diffusion process for attribute synthesis and a graph generator based on the Stochastic Block Model for structure generation.
  • Experiments show that RelDiff outperforms prior methods in generating realistic and coherent synthetic relational databases.

Read Full Article

like

19 Likes

For uninterrupted reading, download the app