UTOPIA: Unlearnable Tabular Data via Decoupled Shortcut Embedding

2026-02-07

AI Summary

Existing unlearning methods exhibit limited efficacy on sensitive tabular data, such as those in finance and healthcare. This work proposes the first decoupled unlearning mechanism tailored to the intrinsic characteristics of tabular data. By leveraging feature saliency analysis, it disentangles high-saliency semantic features from low-saliency redundant ones—using the former for semantic obfuscation and embedding the latter with hyper-correlated shortcuts. Guided by spectral dominance theory, the method constructs constraint-aware dominant shortcuts to achieve certified unlearnability. Extensive experiments across multiple tabular datasets and model architectures demonstrate that the proposed approach reduces unauthorized model performance to near-random levels, substantially outperforming existing techniques while maintaining strong cross-architecture transferability.

Technology Category

• Machine Learning: Unsupervised & Self-Supervised Learning
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories
• Natural Language Processing: Safety and Robustness

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Data transparency and provenance
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Unlearnable examples (UE) have emerged as a practical mechanism to prevent unauthorized model training on private vision data, while extending this protection to tabular data is nontrivial. Tabular data in finance and healthcare is highly sensitive, yet existing UE methods transfer poorly because tabular features mix numerical and categorical constraints and exhibit saliency sparsity, with learning dominated by a few dimensions. Under a Spectral Dominance condition, we show certified unlearnability is feasible when the poison spectrum overwhelms the clean semantic spectrum. Guided by this, we propose Unlearnable Tabular Data via DecOuPled Shortcut EmbeddIng (UTOPIA), which exploits feature redundancy to decouple optimization into two channels: high saliency features for semantic obfuscation and low saliency redundant features for embedding a hyper correlated shortcut, yielding constraint-aware dominant shortcuts while preserving tabular validity. Extensive experiments across tabular datasets and models show UTOPIA drives unauthorized training toward near random performance, outperforming strong UE baselines and transferring well across architectures.

Problem

• unlearnable examples
• tabular data
• data privacy
• spectral dominance
• feature constraints

Innovation

• unlearnable examples
• tabular data
• decoupled optimization
• spectral dominance
• shortcut embedding