INSIGHTS: Demonstration-Based Summaries of Time Series Predictors

TL;DR

INSIGHTS is a new, user-friendly method for providing global explanations of time series models by generating diverse, informative sample summaries that improve understanding and transparency.

Contribution

It introduces a model-agnostic, sample-based approach for global explanation of time series models, emphasizing simplicity, efficiency, and stakeholder interpretability.

Findings

INSIGHTS effectively creates diverse, comprehensive time series summaries.

Domain experts prefer INSIGHTS for understanding model behavior.

User studies show improved understanding with INSIGHTS summaries.

Abstract

Explainability methods have progressed rapidly, but global explanations for time-series models remain underdeveloped, with most approaches focusing on local, instance-level attributions. We introduce INSIGHTS, a model-agnostic, user-centric approach for providing global explanations of time series models. Our approach prioritizes simplicity, efficiency, and transparency in its design, ensuring that stakeholders can readily adopt its outputs. While current methods focus on local explanations, INSIGHTS generates sample summaries that offer a comprehensive overview of model behavior. It balances the importance and diversity of time series samples to create informative subsets using utility functions that capture domain-specific aspects of time series behavior, such as exceeding domain norms. We evaluate INSIGHTS through experiments, interviews, and a user study. Our results indicate INSIGHTS effectively constructs comprehensive, diverse time series subsets, producing summaries manageable for individual evaluation. It is preferred by domain experts for its ability to provide a stable understanding of model behavior and the quality of the samples identified. Moreover, user study participants presented with INSIGHTS-based summaries exhibit an enhanced understanding of the model's overall behavior.