Self-Supervision Improves Diffusion Models for Tabular Data Imputation

TL;DR

This paper presents SimpDM, a self-supervised diffusion model designed for tabular data imputation, which improves stability and robustness over existing methods through alignment and data augmentation strategies.

Contribution

The paper introduces SimpDM, a novel self-supervised diffusion model with alignment and data augmentation techniques tailored for robust tabular data imputation.

Findings

SimpDM outperforms state-of-the-art imputation methods in various scenarios.

Self-supervised alignment improves stability of diffusion-based imputation.

State-dependent data augmentation enhances robustness with limited data.

Abstract

The ubiquity of missing data has sparked considerable attention and focus on tabular data imputation methods. Diffusion models, recognized as the cutting-edge technique for data generation, demonstrate significant potential in tabular data imputation tasks. However, in pursuit of diversity, vanilla diffusion models often exhibit sensitivity to initialized noises, which hinders the models from generating stable and accurate imputation results. Additionally, the sparsity inherent in tabular data poses challenges for diffusion models in accurately modeling the data manifold, impacting the robustness of these models for data imputation. To tackle these challenges, this paper introduces an advanced diffusion model named Self-supervised imputation Diffusion Model (SimpDM for brevity), specifically tailored for tabular data imputation tasks. To mitigate sensitivity to noise, we introduce a self-supervised alignment mechanism that aims to regularize the model, ensuring consistent and stable imputation predictions. Furthermore, we introduce a carefully devised state-dependent data augmentation strategy within SimpDM, enhancing the robustness of the diffusion model when dealing with limited data. Extensive experiments demonstrate that SimpDM matches or outperforms state-of-the-art imputation methods across various scenarios.