Research map / Time series / Long-term forecasting
Predicting the Future by Retrieving the Past
Dazhao Du, Tao Han, Song Guo
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 PFRP global retrieval | SparseTSF · Traffic | MSE: 0.2404 -> 0.1919 | Improved |
| add PFRP global retrieval | SparseTSF · Electricity | MSE: 0.4968 -> 0.3561 | Improved |
| add PFRP global retrieval | DLinear · Weather | MSE: 0.7250 -> 0.6751 | Improved |
| add PFRP global retrieval | SparseTSF, DLinear, PatchTST, TimesNet · ETT datasets | MSE: average gains 3.4%, 3.1%, 1.6%, 2.2% on ETTh1/h2/m1/m2 | Improved |
| add confidence gate (ablation) | PFRP · Electricity | MSE: 0.3960 -> 0.3561 | Improved |
| add output gate (ablation) | PFRP · Electricity | MSE: 0.5140 -> 0.3561 | Improved |
| add local prediction model to global-only PFRP (ablation) | PFRP · Traffic | MSE: 0.1686 -> 0.1919 | Hurt |
| add local prediction model to global-only PFRP (ablation) | PFRP · Electricity | MSE: 0.3952 -> 0.3561 | Improved |
| use feature cosine similarity instead of DTW retrieval (ablation) | PFRP · Electricity | MSE: 0.3659 -> 0.3561 | Improved |
| train encoder with predictive contrastive learning instead of standard contrastive learning (ablation) | PFRP · Electricity | MSE: 0.3833 -> 0.3561 | Improved |
| swap MLP encoder for PatchTST or TimesNet (ablation) | PFRP · Traffic/Electricity | MSE: Traffic TimesNet 0.1754 vs MLP 0.1919; Electricity PatchTST 0.3492 | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- SparseTSF — SparseTSF: Modeling Long-term Time Series Forecasting with *1k* Parameters (ICML 2024) · better on 7 of 7 shared results
- TimesNet — TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysis (ICLR 2023) · better on 2 of 2 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
- Retrieval Augmented Time Series Forecasting (ICML 2025 · similarity 0.87)
- SynTSBench: Rethinking Temporal Pattern Learning in Deep Learning Models for Time Series (NeurIPS 2025 · similarity 0.85)
- ProbTS: Benchmarking Point and Distributional Forecasting across Diverse Prediction Horizons (NeurIPS 2024 · similarity 0.85)
- Towards a General Time Series Forecasting Model with Unified Representation and Adaptive Transfer (ICML 2025 · similarity 0.85)
- Time-LLM: Time Series Forecasting by Reprogramming Large Language Models (ICLR 2024 · similarity 0.85)
- TimeBase: The Power of Minimalism in Efficient Long-term Time Series Forecasting (ICML 2025 · 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.
