LLM-Augmented Bayesian Networks Enhance Few-Shot Tabular Generation

Researchers have developed LAB-Tab, a novel framework for generating tabular data in few-shot scenarios. This method employs a Large Language Model (LLM) to augment Bayesian networks (BNs) by proposing new edges based on source data. A Proximal Policy Optimization (PPO) policy then refines these edges to better align with target domain distributions and downstream utility. In experiments using US Census Bureau data, LAB-Tab demonstrated superior performance in tabular generation with limited target data. AI

IMPACT This framework could improve data synthesis for applications with scarce target datasets, potentially aiding analysis and decision-making.

RANK_REASON The cluster contains an academic paper detailing a new method for tabular data generation. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• Bayesian network
• LAB-Tab
• LLM
• PPO policy
• United States Census Bureau

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →