TabKDE: Simple and Scalable Tabular Data Generation with Kernel Density Estimates

TL;DR

TabKDE introduces a scalable, efficient method for generating synthetic tabular data using kernel density estimates, achieving high accuracy with minimal training time and storage requirements.

Contribution

It presents a novel approach combining copula transformations and kernel density estimates for tabular data generation, enabling scalability and model compactness.

Findings

Achieves comparable accuracy to state-of-the-art methods

Requires almost no training time

Stores models as data coresets, reducing space

Abstract

Tabular data generation considers a large table with multiple columns -- each column comprised of numerical, categorical, or sometimes ordinal values. The goal is to produce new rows for the table that replicate the distribution of rows from the original data -- without just copying those initial rows. The last 4 years have seen enormous progress on this problem, mostly using computational expensive methods that employ one-hot encoding, VAEs, and diffusion. This paper describes a new approach to the problem of tabular data generation. By employing copula transformations and modeling the distribution as a kernel density estimate we can nearly match the accuracy and leakage-avoidance achievements of the previous methods, but with almost no training time. Our method is very scalable, and can be run on data sets orders of magnitude larger than prior state-of-the-art on a simple laptop. Moreover, because we employ kernel density estimates, we can store the model as a coreset of the original data -- we believe the first for generative modeling -- and as a result, require significantly less space as well. Our code is available here: \url{https://github.com/tabkde/tabkde-main}