Perfect Counterfactuals in Imperfect Worlds: Modelling Noisy Implementation of Actions in Sequential Algorithmic Recourse

2024-10-03

Machine-mediated learning

AI Summary

Existing algorithmic recourse methods assume instantaneous, single-step implementation under uniform noise—overlooking the realistic scenario where noise accumulates across sequential actions and is modulated by local data geometry. To address this, we formulate recourse generation as a Markovian sequential decision-making problem—the first such formalization in the literature—and propose ROSE (Robust Sequential Counterfactual Generator). ROSE integrates local-geometry-aware noise modeling with sparsity-inducing, low-cost constraints to ensure action feasibility while substantially improving robustness against perturbation accumulation. Experiments demonstrate that ROSE achieves over 1.8× higher target attainment success rate compared to state-of-the-art baselines, while maintaining computational efficiency and generating sparse, interpretable recourse sequences.

Technology Category

• Reasoning under Uncertainty: Sequential Decision Making
• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• 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

Abstract

Algorithmic recourse suggests actions to individuals who have been adversely affected by automated decision-making, helping them to achieve the desired outcome. Knowing the recourse, however, does not guarantee that users can implement it perfectly, either due to environmental variability or personal choices. Recourse generation should thus anticipate its sub-optimal or noisy implementation. While several approaches construct recourse that is robust to small perturbations – e.g., arising due to its noisy implementation – they assume that the entire recourse is implemented in a single step, thus model the noise as one-off and uniform. But these assumptions are unrealistic since recourse often entails multiple sequential steps, which makes it harder to implement and subject to increasing noise. In this work, we consider recourse under plausible noise that adheres to the local data geometry and accumulates at every step of the way. We frame this problem as a Markov Decision Process and demonstrate that such a distribution of plausible noise satisfies the Markov property. We then propose the RObust SEquential (ROSE) recourse generator for tabular data; our method produces a series of steps leading to the desired outcome even when they are implemented imperfectly. Given plausible modelling of sub-optimal human actions and greater recourse robustness to accumulated uncertainty, ROSE provides users with a high chance of success while maintaining low recourse cost. Empirical evaluation shows that our algorithm effectively navigates the inherent trade-off between recourse robustness and cost while ensuring its sparsity and computational efficiency.

Problem

• Modeling noisy sequential actions in algorithmic recourse
• Ensuring robust recourse under accumulating implementation noise
• Balancing recourse robustness, cost, and sparsity effectively

Innovation

• Models sequential noisy actions using Markov Decision Process
• Generates robust recourse with ROSE for tabular data
• Ensures high success rate with low recourse cost