Generating Sparse Counterfactual Explanations For Multivariate Time Series

TL;DR

This paper introduces SPARCE, a GAN-based method for generating sparse counterfactual explanations for multivariate time series, improving interpretability by focusing on salient features with minimal modifications.

Contribution

The paper presents a novel GAN architecture with a custom sparsity layer for generating sparse counterfactuals in multivariate time series, emphasizing saliency and smoothness.

Findings

Achieves sparser modifications than existing methods

Performs comparably or better on interpretability metrics

Modifies mainly salient time steps and features

Abstract

Since neural networks play an increasingly important role in critical sectors, explaining network predictions has become a key research topic. Counterfactual explanations can help to understand why classifier models decide for particular class assignments and, moreover, how the respective input samples would have to be modified such that the class prediction changes. Previous approaches mainly focus on image and tabular data. In this work we propose SPARCE, a generative adversarial network (GAN) architecture that generates SPARse Counterfactual Explanations for multivariate time series. Our approach provides a custom sparsity layer and regularizes the counterfactual loss function in terms of similarity, sparsity, and smoothness of trajectories. We evaluate our approach on real-world human motion datasets as well as a synthetic time series interpretability benchmark. Although we make significantly sparser modifications than other approaches, we achieve comparable or better performance on all metrics. Moreover, we demonstrate that our approach predominantly modifies salient time steps and features, leaving non-salient inputs untouched.