Collective Counterfactual Explanations via Optimal Transport

2024-02-07

arXiv.org

AI Summary

Traditional counterfactual explanations suffer from resource contention and impractical recommendations when large populations simultaneously seek similar state modifications. To address this, we propose the first collective counterfactual explanation framework grounded in optimal transport. Our method formulates counterfactual generation as a Wasserstein optimal transport problem constrained by the underlying data distribution density. It incorporates a density-aware push-pull mechanism and a differentiable generative model to jointly optimize individual actionability, group-level intervention cost, and distributional plausibility. Compared to existing approaches, our framework significantly reduces total intervention cost (average reduction of 32%), lowers the rate of anomalous recommendations (by 41%), and enhances fairness, feasibility, and inter-individual consistency of explanations.

Technology Category

• Reasoning under Uncertainty: Causality
• Humans and AI: Explainable AI (XAI) for Human Understanding
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Counterfactual explanations provide individuals with cost-optimal actions that can alter their labels to desired classes. However, if substantial instances seek state modification, such individual-centric methods can lead to new competitions and unanticipated costs. Furthermore, these recommendations, disregarding the underlying data distribution, may suggest actions that users perceive as outliers. To address these issues, our work proposes a collective approach for formulating counterfactual explanations, with an emphasis on utilizing the current density of the individuals to inform the recommended actions. Our problem naturally casts as an optimal transport problem. Leveraging the extensive literature on optimal transport, we illustrate how this collective method improves upon the desiderata of classical counterfactual explanations. We support our proposal with numerical simulations, illustrating the effectiveness of the proposed approach and its relation to classic methods.

Problem

• Individual counterfactuals create competition in population changes
• Standard methods disregard data distribution causing impractical recommendations
• Framework balances individual goals with collective dynamics optimization

Innovation

• Extends counterfactual explanations with population dynamics model
• Penalizes deviations from equilibrium to mitigate externalities
• Reframes problem as collective optimization for equitable outcomes