Research map / Time series / Causal and generative time series
Exposing Vulnerabilities in Explanation for Time Series Classifiers via Dual-Target Attacks
Bohan Wang, Zewen Liu, Lu Lin, Hui Liu, Li Xiong, Ming Jin, Wei Jin
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add temporal vulnerability mask Mt (ablation) | TSEF with TimeX++ · SeqComb-MV | ASR: 0.491 -> 0.789 | Improved |
| add frequency perturbation filter Mf (ablation) | TSEF with Integrated Gradients · SeqComb-MV | AUPRC: 0.253 -> 0.334 | Improved |
| add temporal vulnerability mask Mt (ablation) | TSEF with Integrated Gradients · SeqComb-MV | ASR: 0.544 -> 0.781 | Improved |
| add frequency perturbation filter Mf (ablation) | TSEF with TimeX++ · SeqComb-MV | ASR: 0.678 -> 0.789 | Improved |
| replace PGD with dual-target TSEF | TimeX++ explainer · LowVar | AUPRC: 0.258 -> 0.845 | Improved |
| replace PGD with TSEF | TimeX++ explainer · SeqComb-MV | AUPRC: 0.484 -> 0.661 | Improved |
| replace PGD with TSEF | Integrated Gradients · SeqComb-MV | AUPRC: 0.183 -> 0.334 | Improved |
| replace PGD with TSEF | TimeX explainer · ECG | AUPRC: 0.320 -> 0.419 | Improved |
| replace ADV2 with TSEF | TimeX++ explainer · PAM | AUPRC: 0.820 -> 0.876 | Improved |
| replace ADV2 and PGD with TSEF | TimeX explainer · Epilepsy | AUPRC: 0.922 -> 0.981 | Improved |
| replace PGD with TSEF | TimeX++ explainer · ECG | ASR: 0.946 -> 0.951 | No clear effect |
| replace PGD with TSEF | Integrated Gradients · Epilepsy | ASR: 0.191 -> 0.192 | No clear effect |
Closest papers by meaning
- TimeX++: Learning Time-Series Explanations with Information Bottleneck (ICML 2024 · similarity 0.87)
- Optimal Information Retention for Time-Series Explanations (ICML 2025 · similarity 0.86)
- InteDisUX: Intepretation-Guided Discriminative User-Centric Explanation for Time Series (AAAI 2025 · similarity 0.85)
- Inherently Interpretable Time Series Classification via Multiple Instance Learning (ICLR 2024 · similarity 0.84)
- Explaining Time Series via Contrastive and Locally Sparse Perturbations (ICLR 2024 · similarity 0.83)
- TimeXL: Explainable Multi-modal Time Series Prediction with LLM-in-the-Loop (NeurIPS 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
