P$^2$CE: Model-Agnostic Plausible Pareto-Optimal Counterfactual Explanations

2026-06-16

AI Summary

Existing counterfactual explanation methods struggle to simultaneously achieve feasibility, plausibility, and computational efficiency, limiting their applicability in high-stakes decision-making scenarios. This work proposes P²CE, a model-agnostic algorithm that uniquely integrates Isolation Forests with SHAP values to enable efficient counterfactual generation. By leveraging Isolation Forests to ensure generated counterfactuals adhere to the underlying data distribution and employing SHAP values to accelerate the optimization process, P²CE produces a Pareto-optimal set of solutions across multiple feasibility dimensions, offering diverse and actionable recommendations. Experimental results on three benchmark datasets demonstrate that P²CE significantly outperforms state-of-the-art methods, achieving superior solution quality and computational efficiency while maintaining high plausibility and diversity.

Abstract

The increasing use of machine learning algorithms in social applications has raised concerns about fairness and transparency, leading to the development of counterfactual explanations. These explanations supports individuals to understand and potentially alter unfavorable decisions in areas such as loan applications, job selections, and more, by providing actionable changes to input features that would lead to a desired outcome. Existing methods often struggle to balance feasibility, plausibility, and computational efficiency. To address this, we introduce P$^2$CE, an algorithm for generating plausible Pareto-optimal counterfactual explanations, offering users a diverse set of optimal trade-offs between different notions of feasibility. P$^2$CE employs an auxiliary isolation forest outlier detector to ensure that explanations are in accordance with the data distribution and leverages SHAP values to obtain optimal results with short computing times, regardless of the underlying model. Our algorithm was empirically evaluated on three datasets, demonstrating superior performance in terms of both solution quality and computational efficiency compared to related techniques.

Problem

• counterfactual explanations
• feasibility
• plausibility
• computational efficiency
• Pareto-optimal

Innovation

• counterfactual explanations
• Pareto optimality
• model-agnostic
• isolation forest
• SHAP values