Research map / Time series / Long-term forecasting
T3Time: Tri-Modal Time Series Forecasting via Adaptive Multi-Head Alignment and Residual Fusion
Abdul Monaf Chowdhury, Rabeya Akter, Safaeid Hossain Arib
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 channel-wise residual fusion (ablation) | T3Time · ILI (avg horizons) | MSE: 2.176 -> 1.705 | Improved |
| add frequency encoding branch (ablation) | T3Time · 7 benchmarks (avg) | MSE: 3.22% MSE and 1.85% MAE drop when removed | Improved |
| add adaptive multi-head cross-modal alignment (ablation) | T3Time · ILI (avg horizons) | MSE: 1.813 -> 1.705 | Improved |
| add horizon-aware gating (ablation) | T3Time · ILI (avg horizons) | MSE: 1.724 -> 1.705 | Improved |
| add horizon-aware gating (ablation) | T3Time · ETTm1 (avg horizons) | MSE: 0.373 -> 0.372 | No clear effect |
| add frequency branch (ablation) | T3Time · ETTm2 (avg horizons) | MSE: 0.279 -> 0.279 | No clear effect |
| add adaptive multi-head cross-modal alignment (ablation) | T3Time · ETTh2 (avg horizons) | MSE: 0.355 -> 0.348 | Hurt |
| replace TimeCMA with T3Time (tri-modal) | TimeCMA · ETTm2 (avg horizons) | MSE: 0.275 -> 0.279 | Hurt |
| replace TimeCMA with T3Time | TimeCMA · ILI (avg horizons) | MSE: 1.922 -> 1.705 | Improved |
| replace TimeCMA with T3Time | TimeCMA · few-shot 10% data | MSE: -7.13% average MSE | Improved |
| use T3Time instead of best SOTA baselines | SOTA baselines · few-shot 5% data | MSE and MAE: 4.13% MSE and 1.91% MAE | Improved |
| use T3Time instead of best SOTA baselines | SOTA baselines · 8 long-term benchmarks | MSE and MAE: 3.28% MSE and 2.29% MAE | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- TimeCMA — TimeCMA: Towards LLM-Empowered Multivariate Time Series Forecasting via Cross-Modality Alignment (AAAI 2025) · better on 10 of 11 shared results
- iTransformer — iTransformer: Inverted Transformers Are Effective for Time Series Forecasting (ICLR 2024) · better on 2 of 2 shared results
- Time-LLM — Time-LLM: Time Series Forecasting by Reprogramming Large Language Models (ICLR 2024) · better on 1 of 1 shared result
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
- TimeCMA: Towards LLM-Empowered Multivariate Time Series Forecasting via Cross-Modality Alignment (AAAI 2025 · similarity 0.90)
- TimeXL: Explainable Multi-modal Time Series Prediction with LLM-in-the-Loop (NeurIPS 2025 · similarity 0.90)
- CALF: Aligning LLMs for Time Series Forecasting via Cross-modal Fine-Tuning (AAAI 2025 · similarity 0.89)
- Unlocking the Power of Patch: Patch-Based MLP for Long-Term Time Series Forecasting (AAAI 2025 · similarity 0.89)
- Learning Pattern-Specific Experts for Time Series Forecasting Under Patch-level Distribution Shift (NeurIPS 2025 · similarity 0.89)
- ARM: Refining Multivariate Forecasting with Adaptive Temporal-Contextual Learning (ICLR 2024 · similarity 0.89)
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.
