Temporally Unified Adversarial Perturbations for Time Series Forecasting

2026-02-12

AI Summary

This work addresses a critical limitation in existing time series adversarial attacks: the neglect of temporal consistency, which leads to conflicting perturbations at the same timestamp across overlapping samples and hinders real-world deployment. To resolve this, the authors propose Temporally Unified Adversarial Perturbations (TUAPs), which introduce, for the first time, a temporal consistency constraint to ensure identical perturbations at shared timestamps within overlapping windows. They further design a modular Timestamp-level Gradient Accumulation Method (TGAM) that integrates a momentum mechanism to aggregate local gradients, enabling efficient exploration of the perturbation space. Extensive experiments demonstrate that TUAPs significantly outperform current methods across three benchmark datasets and four state-of-the-art models, achieving superior performance in white-box, black-box transfer, and unconstrained attack scenarios.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Time-Series/Data Streams
• Multiagent Systems: Adversarial Agents

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

While deep learning models have achieved remarkable success in time series forecasting, their vulnerability to adversarial examples remains a critical security concern. However, existing attack methods in the forecasting field typically ignore the temporal consistency inherent in time series data, leading to divergent and contradictory perturbation values for the same timestamp across overlapping samples. This temporally inconsistent perturbations problem renders adversarial attacks impractical for real-world data manipulation. To address this, we introduce Temporally Unified Adversarial Perturbations (TUAPs), which enforce a temporal unification constraint to ensure identical perturbations for each timestamp across all overlapping samples. Moreover, we propose a novel Timestamp-wise Gradient Accumulation Method (TGAM) that provides a modular and efficient approach to effectively generate TUAPs by aggregating local gradient information from overlapping samples. By integrating TGAM with momentum-based attack algorithms, we ensure strict temporal consistency while fully utilizing series-level gradient information to explore the adversarial perturbation space. Comprehensive experiments on three benchmark datasets and four representative state-of-the-art models demonstrate that our proposed method significantly outperforms baselines in both white-box and black-box transfer attack scenarios under TUAP constraints. Moreover, our method also exhibits superior transfer attack performance even without TUAP constraints, demonstrating its effectiveness and superiority in generating adversarial perturbations for time series forecasting models.

Problem

• adversarial perturbations
• time series forecasting
• temporal consistency
• overlapping samples

Innovation

• Temporally Unified Adversarial Perturbations
• Time Series Forecasting
• Adversarial Attacks
• Temporal Consistency
• Gradient Accumulation