Multi-Objective Counterfactual Explanations

By Susanne Dandl, Christoph Molnar, Martin Binder, Bernd Bischl. Parallel Problem Solving from Nature, 2020

322 citations recorded by the source corpus.

Counterfactual explanations are one of the most popular methods to make predictions of black box machine learning models interpretable by providing explanations in the form of `what-if scenarios'. Most current approaches optimize a collapsed, weighted sum of multiple objectives, which are naturally difficult to balance a-priori. We propose the Multi-Objective Counterfactuals (MOC) method, which translates the counterfactual search into a multi-objective optimization problem. Our approach not only returns a diverse set of counterfactuals with different trade-offs between the proposed objectives, but also maintains diversity in feature space. This enables a more detailed post-hoc analysis to facilitate better understanding and also more options for actionable user responses to change the predicted outcome. Our approach is also model-agnostic and works for numerical and categorical input features. We show the usefulness of MOC in concrete cases and compare our approach with state-of-the-art methods for counterfactual explanations.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/48ace0d03aa14792922faf4a4a256385b023972c

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

A survey of algorithmic recourse: definitions, formulations, solutions, and prospects (2020, background): “…(2020.03) ViCE [68] E heuristic query (2020.03) Plausible CFs [22] E grad opt + gen alg dataset (2020.04) SEDC-T [175] E heuristic query (2020.04) MOC [49] E gen alg query (2020.04) SCOUT [179] E grad opt gradient (2020.04) ASP-based CFs [28] E answer-set prog. query (2020.05) CBR-based CFs [95] E…”
A Survey of Contrastive and Counterfactual Explanation Generation Methods for Explainable Artificial Intelligence (2021, methodology): “use the hypervolume indicator metric to estimate the quality of the estimated Pareto front during counterfactual search [134].”
Data-centric Artificial Intelligence: A Survey (2023, unknown): “White-box methods necessitate access to the evaluated models, which can be achieved through complete internals [36, 111, 138], gradients [225], or solely the prediction function [52, 57, 127, 202].”
Interpretable Machine Learning - A Brief History, State-of-the-Art and Challenges (2020, methodology): “Popular local IML methods are Shapley values [69,112] and counterfactual explanations [122,20,81,116,118].”
Explaining NLP Models via Minimal Contrastive Editing (MiCE) (2020, background): “While efficient search for targeted edits is an open challenge in other fields of machine learning (Russell, 2019; Dandl et al., 2020), this problem is even more challenging for language data, as the space of possible perturbations is much larger than for tabular data.”
Counterfactual Explanations for Machine Learning: A Review (2020, methodology): “Black-box approaches use gradient-free optimization algorithms such as Nelder-Mead [56], growing spheres [74], FISTA [30, 108], or genetic algorithms [72, 98, 26] to solve the optimization problem.”
Counterfactual Explanations and Algorithmic Recourses for Machine Learning: A Review (2020, unknown): “L0 norm of counterfactual and the input datapoint [67, 191].”