Structure-Aware Robust Counterfactual Explanations via Conditional Gaussian Network Classifiers

2026-02-08

AI Summary

This work addresses the limitation of existing counterfactual explanation methods that neglect structural dependencies among variables, often yielding infeasible recommendations. To overcome this, the authors propose a novel framework based on Conditional Gaussian Network Classifiers (CGNCs), which uniquely integrates the directed acyclic graph structure of CGNCs directly into the counterfactual generation process. This integration inherently ensures conditional consistency among features without requiring additional constraints. Furthermore, by combining McCormick piecewise relaxations with cutting-plane optimization, the method reformulates the non-convex quadratic problem into a tractable mixed-integer linear program (MILP), guaranteeing convergence to a globally robust solution. Experiments across multiple datasets demonstrate that the approach efficiently produces structurally consistent and highly robust counterfactual explanations, establishing a scalable optimization paradigm for counterfactual reasoning under non-convex constraints.

Technology Category

• Search and Optimization: Non-convex Optimization
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories
• Reasoning under Uncertainty: Stochastic Optimization

Application Category

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

Abstract

Counterfactual explanation (CE) is a core technique in explainable artificial intelligence (XAI), widely used to interpret model decisions and suggest actionable alternatives. This work presents a structure-aware and robustness-oriented counterfactual search method based on the conditional Gaussian network classifier (CGNC). The CGNC has a generative structure that encodes conditional dependencies and potential causal relations among features through a directed acyclic graph (DAG). This structure naturally embeds feature relationships into the search process, eliminating the need for additional constraints to ensure consistency with the model's structural assumptions. We adopt a convergence-guaranteed cutting-set procedure as an adversarial optimization framework, which iteratively approximates solutions that satisfy global robustness conditions. To address the nonconvex quadratic structure induced by feature dependencies, we apply piecewise McCormick relaxation to reformulate the problem as a mixed-integer linear program (MILP), ensuring global optimality. Experimental results show that our method achieves strong robustness, with direct global optimization of the original formulation providing especially stable and efficient results. The proposed framework is extensible to more complex constraint settings, laying the groundwork for future advances in counterfactual reasoning under nonconvex quadratic formulations.

Problem

• Counterfactual Explanation
• Robustness
• Structure Awareness
• Conditional Dependencies
• Causal Relations

Innovation

• counterfactual explanation
• conditional Gaussian network
• structure-aware
• robust optimization
• mixed-integer linear programming