CF-OPT: Counterfactual Explanations for Structured Prediction

TL;DR

CF-OPT introduces a method to generate plausible counterfactual explanations for structured prediction models using variational autoencoders, enhancing interpretability of complex neural-optimization pipelines.

Contribution

The paper presents CF-OPT, a novel first-order optimization algorithm leveraging VAEs to produce counterfactual explanations in structured learning models.

Findings

Provides close and plausible counterfactual explanations

Works across a broad class of structured learning architectures

Improves interpretability of neural-optimization pipelines

Abstract

Optimization layers in deep neural networks have enjoyed a growing popularity in structured learning, improving the state of the art on a variety of applications. Yet, these pipelines lack interpretability since they are made of two opaque layers: a highly non-linear prediction model, such as a deep neural network, and an optimization layer, which is typically a complex black-box solver. Our goal is to improve the transparency of such methods by providing counterfactual explanations. We build upon variational autoencoders a principled way of obtaining counterfactuals: working in the latent space leads to a natural notion of plausibility of explanations. We finally introduce a variant of the classic loss for VAE training that improves their performance in our specific structured context. These provide the foundations of CF-OPT, a first-order optimization algorithm that can find counterfactual explanations for a broad class of structured learning architectures. Our numerical results show that both close and plausible explanations can be obtained for problems from the recent literature.