Research map / Time series / Long-term forecasting
EasyTS: The Express Lane to Long Time Series Forecasting
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Presents EasyTS, a toolkit bundling datasets (including four new open-source electricity datasets), time series analysis/visualization tools, and a model-building and evaluation interface for long time series forecasting.
Problem
Long-term time series forecasting research is hampered by scattered datasets, tedious data preprocessing and analysis, and time-consuming model building and comparison against existing baselines.
Approach
A unified one-stop toolkit that integrates dataloaders for many LTSF datasets, visualization and analysis tools, and a model verification interface with multiple metrics and built-in baselines (DLinear), lowering the entry barrier for LTSF research.
Key results
- Introduces four new open-source electricity-domain datasets (FOOD, MANU, PHAR, OFFICE) collected from industrial scenarios.
- Toolkit covers datasets from six domains, including energy, transportation, cloud, weather, and disease.
- Provides eight evaluation metrics (MSE, MAE, MAPE, MSPE, RSE, RMSE, CORR, DDTW) and built-in DLinear baseline; no quantitative model results reported.
Limitations
- No quantitative benchmark experiments comparing models are reported.
- Built-in baseline set is limited to DLinear in the described version.
- Target user is primarily novices; advanced research workflow support not evaluated.
Experimental setup
Datasets: ECL ETTm Solar Wind APP Flow ECW Weather ILI Traffic METR-LA PEMS-BAY FOOD
Metrics: MSE MAE MAPE MSPE RSE RMSE CORR DDTW
Baselines: DLinear
Builds on: AllenNLP
Closest papers by meaning
- DeformableTST: Transformer for Time Series Forecasting without Over-reliance on Patching (NeurIPS 2024 · similarity 0.84)
- TimeBase: The Power of Minimalism in Efficient Long-term Time Series Forecasting (ICML 2025 · similarity 0.83)
- A Closer Look at Transformers for Time Series Forecasting: Understanding Why They Work and Where They Struggle (ICML 2025 · similarity 0.82)
- Cherry-Picking in Time Series Forecasting: How to Select Datasets to Make Your Model Shine (AAAI 2025 · similarity 0.82)
- VerbalTS: Generating Time Series from Texts (ICML 2025 · similarity 0.82)
- LightGTS: A Lightweight General Time Series Forecasting Model (ICML 2025 · similarity 0.82)
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.