Research map / Time series / Long-term forecasting
Decentralized Attention Fails Centralized Signals: Rethinking Transformers for Medical Time Series
Guoqi Yu, Juncheng Wang, Chen Yang, Jing Qin, Angelica I Aviles-Rivero, Shujun Wang
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 dual (temporal plus channel) tokenization instead of no tokenization (ablation) | TeCh · APAVA | accuracy: 50.68 -> 86.86 | Improved |
| add dual tokenization instead of temporal-only (ablation) | TeCh · APAVA | accuracy: 55.93 -> 86.86 | Improved |
| add dual tokenization instead of channel-only (ablation) | TeCh · PTB | accuracy: 85.96 -> 84.15 | Hurt |
| add dual tokenization instead of temporal-only (ablation) | TeCh · TDBrain | accuracy: 93.21 -> 89.79 | Hurt |
| add dual tokenization instead of channel-only (ablation) | TeCh · UCI-HAR | accuracy: 92.98 -> 94.15 | Improved |
| add CoTAR token interaction instead of removing it (ablation) | TeCh · APAVA | accuracy: 83.31 -> 86.86 | Improved |
| replace attention with CoTAR (ablation) | TeCh · TDBrain | accuracy: 90.40 -> 93.21 | Improved |
| replace attention with CoTAR (ablation) | TeCh · UCI-HAR | accuracy: 93.13 -> 94.15 | Improved |
| replace attention with CoTAR (ablation) | TeCh · PTB | accuracy: 85.74 -> 85.96 | Improved |
| add CoTAR instead of removing it (ablation) | TeCh · ADFTD | F1: 47.26 -> 48.84 | No clear effect |
| replace Medformer with TeCh | TeCh · ADFTD | average of six metrics: 54.63 -> 54.16 | Hurt |
| replace Medformer with TeCh | TeCh · APAVA | average of six metrics: 79.74 -> 88.99 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Medformer — Medformer: A Multi-Granularity Patching Transformer for Medical Time-Series Classification (NeurIPS 2024) · better on 13 of 14 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Medformer: A Multi-Granularity Patching Transformer for Medical Time-Series Classification (NeurIPS 2024 · similarity 0.89)
- MAESTRO : Adaptive Sparse Attention and Robust Learning for Multimodal Dynamic Time Series (NeurIPS 2025 · similarity 0.84)
- GEM: Empowering MLLM for Grounded ECG Understanding with Time Series and Images (NeurIPS 2025 · similarity 0.84)
- NeuroAtlas: Benchmarking Foundation Models for Clinical EEG and Brain-Computer Interfaces (NeurIPS 2026 · similarity 0.84)
- Revitalizing Medical Time Series with Vision-Informed Retrieval: A Vision-Language Perspective (NeurIPS 2026 · similarity 0.83)
- CBraMod: A Criss-Cross Brain Foundation Model for EEG Decoding (ICLR 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.
