Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

DISCO-TAB: A Hierarchical Reinforcement Learning Framework for Privacy-Preserving Synthesis of Complex Clinical Data

Arshia Ilaty, Hossein Shirazi, Amir Rahmani, Hajar Homayouni

Abstract

The development of robust clinical decision support systems is frequently impeded by the scarcity of high-fidelity, privacy-preserving biomedical data. While Generative Large Language Models (LLMs) offer a promising avenue for synthetic data generation, they often struggle to capture the complex, non-linear dependencies and severe class imbalances inherent in Electronic Health Records (EHR), leading to statistically plausible but clinically invalid records. To bridge this gap, we introduce DISCO-TAB (DIScriminator-guided COntrol for TABular synthesis), a novel framework that orchestrates a fine-tuned LLM with a multi-objective discriminator system optimized via Reinforcement Learning. Unlike prior methods relying on scalar feedback, DISCO-TAB evaluates synthesis at four granularities, token, sentence, feature, and row, while integrating Automated Constraint Discovery and Inverse-Frequency Reward Shaping to autonomously preserve latent medical logic and resolve minority-class collapse. We rigorously validate our framework across diverse benchmarks, including high-dimensional, small-sample medical datasets (e.g., Heart Failure, Parkinson's). Our results demonstrate that hierarchical feedback yields state-of-the-art performance, achieving up to 38.2% improvement in downstream clinical classifier utility compared to GAN and Diffusion baselines, while ensuring exceptional statistical fidelity (JSD < 0.01) and robust resistance to membership inference attacks. This work establishes a new standard for generating trustworthy, utility-preserving synthetic tabular data for sensitive healthcare applications.

DISCO-TAB: A Hierarchical Reinforcement Learning Framework for Privacy-Preserving Synthesis of Complex Clinical Data

Abstract

The development of robust clinical decision support systems is frequently impeded by the scarcity of high-fidelity, privacy-preserving biomedical data. While Generative Large Language Models (LLMs) offer a promising avenue for synthetic data generation, they often struggle to capture the complex, non-linear dependencies and severe class imbalances inherent in Electronic Health Records (EHR), leading to statistically plausible but clinically invalid records. To bridge this gap, we introduce DISCO-TAB (DIScriminator-guided COntrol for TABular synthesis), a novel framework that orchestrates a fine-tuned LLM with a multi-objective discriminator system optimized via Reinforcement Learning. Unlike prior methods relying on scalar feedback, DISCO-TAB evaluates synthesis at four granularities, token, sentence, feature, and row, while integrating Automated Constraint Discovery and Inverse-Frequency Reward Shaping to autonomously preserve latent medical logic and resolve minority-class collapse. We rigorously validate our framework across diverse benchmarks, including high-dimensional, small-sample medical datasets (e.g., Heart Failure, Parkinson's). Our results demonstrate that hierarchical feedback yields state-of-the-art performance, achieving up to 38.2% improvement in downstream clinical classifier utility compared to GAN and Diffusion baselines, while ensuring exceptional statistical fidelity (JSD < 0.01) and robust resistance to membership inference attacks. This work establishes a new standard for generating trustworthy, utility-preserving synthetic tabular data for sensitive healthcare applications.

Paper Structure

This paper contains 52 sections, 20 equations, 1 figure, 7 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Adversarial and Auto-Encoder Methods
  4. Diffusion Models for Tabular Synthesis
  5. Autoregressive Transformers and LLM-Based Baselines
  6. Constraint- and Feedback-Driven Optimization
  7. Privacy Evaluation and Attacks on Synthetic Tabular Releases
  8. Methodology
  9. Problem Formulation and Linearization
  10. Automated Constraint Discovery
  11. Hierarchical Multi-Objective Discriminators
  12. Level 1: Token Consistency ($D_{token}$)
  13. Level 2: Semantic Plausibility ($D_{sent}$)
  14. Level 3: Feature Correlation ($D_{feat}$)
  15. Level 4: Global Coherence ($D_{row}$)
  16. ...and 37 more sections

Figures (1)

  • Figure 1: The DISCO-TAB framework for tabular data synthesis.(A) Automated Constraint Discovery: The system analyzes the real dataset ($\mathcal{D}_{real}$) to compute a global correlation matrix and identify statistically significant feature dependencies (e.g., Age vs. Comorbidities), which are used to initialize the feature-level discriminator ($D_{feat}$). (B) Hierarchical Discriminators: A multi-objective discriminator ensemble evaluates generated records ($S$) at four semantic granularities: token-level (syntactic validity), sentence-level (local semantic plausibility), feature-level (dependency consistency), and row-level (global coherence). (C) Inverse Frequency Reward Shaping (IFRS): During reinforcement learning, the PPO reward is scaled based on the rarity of the target class associated with the generated record, mitigating mode collapse in imbalanced datasets.