Research map / Time series / Long-term forecasting
Time2Graph: Revisiting Time Series Modeling with Dynamic Shapelets
Ziqiang Cheng, Yang Yang, Wei Wang, Wenjie Hu, Yueting Zhuang, Guojie Song
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 time-aware shapelets and their learned timing factors (ablation) | Time2Graph · EQS | accuracy: 76.98 -> 79.14 | Improved |
| Add time-aware shapelets and their learned timing factors (ablation) | Time2Graph · WTC | accuracy: 70.13 -> 72.73 | Improved |
| Add time-aware shapelets and their learned timing factors (ablation) | Time2Graph · STB | accuracy: 95.68 -> 96.76 | Improved |
| Add time-aware shapelets and their learned timing factors (ablation) | Time2Graph · ECR | F1: 31.36 -> 34.44 | Improved |
| Add time-aware shapelets and their learned timing factors (ablation) | Time2Graph · NTF | F1: 42.11 -> 62.97 | Improved |
| Replace weighted shapelet assignments with only the highest-probability shapelet per segment, then use LSTM sequence classification (ablation) | Time2Graph · EQS | accuracy: 79.14 -> 75.53 | Hurt |
| Replace weighted shapelet assignments with only the highest-probability shapelet per segment, then use LSTM sequence classification (ablation) | Time2Graph · WTC | accuracy: 72.73 -> 55.84 | Hurt |
| Replace weighted shapelet assignments with only the highest-probability shapelet per segment, then use LSTM sequence classification (ablation) | Time2Graph · STB | accuracy: 96.76 -> 78.10 | Hurt |
| Replace weighted shapelet assignments with only the highest-probability shapelet per segment, then use LSTM sequence classification (ablation) | Time2Graph · ECR | F1: 34.44 -> 23.66 | Hurt |
| Replace weighted shapelet assignments with only the highest-probability shapelet per segment, then use LSTM sequence classification (ablation) | Time2Graph · NTF | F1: 62.97 -> 28.44 | Hurt |
| Use Time2Graph instead of the strongest listed baseline | Time series classification · WTC | accuracy: 74.42 -> 72.73 | Hurt |
| Use Time2Graph instead of the strongest listed baseline | Time series classification · STB | accuracy: 97.03 -> 96.76 | Hurt |
Closest papers by meaning
- ShapeX: Shapelet-Driven Post Hoc Explanations for Time Series Classification Models (NeurIPS 2025 · similarity 0.90)
- Diffusion Language-Shapelets for Semi-supervised Time-Series Classification (AAAI 2024 · similarity 0.87)
- Learning Soft Sparse Shapes for Efficient Time-Series Classification (ICML 2025 · similarity 0.86)
- Abstracted Shapes as Tokens - A Generalizable and Interpretable Model for Time-series Classification (NeurIPS 2024 · similarity 0.86)
- TimeDART: A Diffusion Autoregressive Transformer for Self-Supervised Time Series Representation (ICML 2025 · similarity 0.85)
- Time-Aware Knowledge Representations of Dynamic Objects with Multidimensional Persistence (AAAI 2024 · similarity 0.85)
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.
