Consistent Sufficient Explanations and Minimal Local Rules for explaining regression and classification models

TL;DR

This paper introduces probabilistic sufficient explanations (P-SE) for models, providing minimal feature subsets that reliably explain predictions, extending to regression, handling non-discrete features, and offering local rule-based explanations with a new estimator and open-source tools.

Contribution

It develops a fast, consistent estimator for P-SE, extends explanations to regression, and introduces local rule-based explanations without needing distributional assumptions.

Findings

The estimator is accurate and efficient for various data types.

P-SE can be extended reliably to regression problems.

The proposed methods outperform existing explainability techniques.

Abstract

To explain the decision of any model, we extend the notion of probabilistic Sufficient Explanations (P-SE). For each instance, this approach selects the minimal subset of features that is sufficient to yield the same prediction with high probability, while removing other features. The crux of P-SE is to compute the conditional probability of maintaining the same prediction. Therefore, we introduce an accurate and fast estimator of this probability via random Forests for any data $(\boldsymbol{X}, Y)$ and show its efficiency through a theoretical analysis of its consistency. As a consequence, we extend the P-SE to regression problems. In addition, we deal with non-discrete features, without learning the distribution of $\boldsymbol{X}$ nor having the model for making predictions. Finally, we introduce local rule-based explanations for regression/classification based on the P-SE and compare our approaches w.r.t other explainable AI methods. These methods are available as a Python package at \url{www.github.com/salimamoukou/acv00}.