Tabular Data Contrastive Learning via Class-Conditioned and Feature-Correlation Based Augmentation

TL;DR

This paper introduces a class-conditioned and feature-correlation-based augmentation method for contrastive learning on tabular data, significantly improving classification performance by more effective data corruption strategies.

Contribution

It proposes a novel augmentation technique for tabular contrastive learning that conditions data corruption on class labels and feature correlations, enhancing model robustness.

Findings

Outperforms traditional augmentation methods in tabular data classification

Improves contrastive learning effectiveness through class-aware data corruption

Demonstrates consistent gains across multiple datasets

Abstract

Contrastive learning is a model pre-training technique by first creating similar views of the original data, and then encouraging the data and its corresponding views to be close in the embedding space. Contrastive learning has witnessed success in image and natural language data, thanks to the domain-specific augmentation techniques that are both intuitive and effective. Nonetheless, in tabular domain, the predominant augmentation technique for creating views is through corrupting tabular entries via swapping values, which is not as sound or effective. We propose a simple yet powerful improvement to this augmentation technique: corrupting tabular data conditioned on class identity. Specifically, when corrupting a specific tabular entry from an anchor row, instead of randomly sampling a value in the same feature column from the entire table uniformly, we only sample from rows that are identified to be within the same class as the anchor row. We assume the semi-supervised learning setting, and adopt the pseudo labeling technique for obtaining class identities over all table rows. We also explore the novel idea of selecting features to be corrupted based on feature correlation structures. Extensive experiments show that the proposed approach consistently outperforms the conventional corruption method for tabular data classification tasks. Our code is available at https://github.com/willtop/Tabular-Class-Conditioned-SSL.