Work

Reward-Guided Generation Improves the Scientific Utility of Synthetic Biomedical Data

Year: 2026

Type: preprint

Abstract: Synthetic data generation is a promising approach for biomedical data sharing and dataset augmentation, yet existing methods lack mechanisms to preserve statistical properties necessary for scientific analysis. To address this, we introduce RLSYN+REG, a reinforcement learning-driven generative model, which encourages that regression models trained on synthetic data reproduce the coefficients and predictions of their real-data counterparts. We evaluate RL-S yn +R eg on MIMIC-III and the American Community Survey (ACS) across regression model reproduction, fidelity to real data, and privacy. Synthetic data from RLS yn +R eg substantially improves upon that of RLS yn , raising correlations between real and synthetic regression coefficients from 0.054 to 0.600 on MIMIC-III and from 0.160 to 0.376 on ACS. Predictive performance also improves, reducing the gap between real-data baselines by 81.4% and 97.6% on MIMIC-III and ACS, respectively. These improvements come with negligible cost to fidelity or privacy and are robust to reductions in training data.

Source: medRxiv

DOI: 10.64898/2026.03.11.26348077

Authors: Nicholas J. Jackson, Natalia Espinosa-Dice, Chao Yan, Bradley A. Malin

Institutions: Vanderbilt University, Princeton University, Vanderbilt University Medical Center

Language: English

Topic: Privacy-Preserving Technologies in Data

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

Funders: National Institutes of Health, U.S. National Library of Medicine

Awards: G2175434712, Vanderbilt Biomedical Informatics Training Program, A FAIR Bridge2AI Center (FABRIC)