Towards Ground Truth Explainability on Tabular Data

TL;DR

This paper introduces a method for creating synthetic tabular datasets with known properties using copulas, enabling ground truth explanations to improve interpretability and validation of explainability methods.

Contribution

It proposes a novel approach to generate synthetic data with controlled statistical properties for ground truth explainability in tabular data.

Findings

Synthetic datasets help validate explainability methods.

Copulas effectively specify desired data properties.

Experiments demonstrate improved understanding of feature impacts.

Abstract

In data science, there is a long history of using synthetic data for method development, feature selection and feature engineering. Our current interest in synthetic data comes from recent work in explainability. Today's datasets are typically larger and more complex - requiring less interpretable models. In the setting of \textit{post hoc} explainability, there is no ground truth for explanations. Inspired by recent work in explaining image classifiers that does provide ground truth, we propose a similar solution for tabular data. Using copulas, a concise specification of the desired statistical properties of a dataset, users can build intuition around explainability using controlled data sets and experimentation. The current capabilities are demonstrated on three use cases: one dimensional logistic regression, impact of correlation from informative features, impact of correlation from redundant variables.