Beyond Sufficiency: Time Series Explanation with Counterfactual Necessity

2026-07-23

AI Summary

This work addresses a critical limitation in existing time series explanation methods, which predominantly emphasize sufficiency and often misidentify non-essential yet supportive subsequences as important, thereby overlooking components truly necessary for model decisions. To remedy this, we propose TimePNS, a novel framework that introduces counterfactual necessity—drawn from Pearl’s causal theory—into time series interpretability for the first time. TimePNS employs a two-stage design: first learning an identifiable causal generative process alongside a sufficiency mask, then evaluating the necessity of temporal factors through counterfactual interventions. By integrating causal generative modeling, temporal gating mechanisms, and interpretable mask learning, our approach significantly outperforms strong baselines on both synthetic and real-world datasets, enabling more accurate identification of decision-critical subsequences and achieving an effective balance between sufficiency and necessity.

Abstract

Faithful explanations of time-series classifiers should identify subsequences that are not only sufficient to preserve a black-box model's prediction, but also necessary for maintaining it. However, existing sufficiency-oriented methods can assign high importance to spurious subsequences that support the prediction without being essential to the model's decision. We introduce \textbf{TimePNS}, a necessity-aware framework for time-series explanation. Inspired by Pearl's counterfactual notion of necessity, TimePNS assesses whether a temporal factor is necessary by intervening on it and measuring whether the original prediction is disrupted. The framework adopts a two-stage design. Stage I learns an identifiable causal generative process together with a sufficiency-oriented explanation mask. Stage II performs counterfactual interventions on temporal factors to derive necessity signals, which supervise a temporal gate that refines the initial explanation by suppressing non-essential components and emphasizing counterfactually necessary ones. Experiments on synthetic and real-world time-series benchmarks show that TimePNS more accurately identifies decision-critical subsequences and consistently improves sufficiency-necessity trade-offs over strong baselines.

Problem

• time series explanation
• counterfactual necessity
• sufficiency
• model interpretability
• temporal subsequence

Innovation

• counterfactual necessity
• time series explanation
• causal generative model
• temporal intervention
• necessity-aware explanation