![[2602.18465] Revisiting the Seasonal Trend Decomposition for Enhanced Time Series Forecasting](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.18465] Revisiting the Seasonal Trend Decomposition for Enhanced Time Series Forecasting
Summary
This paper presents an improved method for multivariate time series forecasting by enhancing seasonal trend decomposition, achieving significant error reduction in predictions.
Why It Matters
Time series forecasting is crucial for various industries, including finance and environmental science. This research offers a novel approach that enhances prediction accuracy, which can lead to better decision-making and resource management in real-world applications.
Key Takeaways
- The study introduces dual-MLP models that improve forecasting accuracy.
- Error rates were reduced by approximately 10% compared to existing models.
- The approach maintains linear time complexity, making it efficient for practical applications.
- The method was validated using benchmark datasets and real-world hydrological data.
- Reversible instance normalization is effective only for trend components, necessitating a different approach for seasonal components.
Computer Science > Machine Learning arXiv:2602.18465 (cs) [Submitted on 6 Feb 2026] Title:Revisiting the Seasonal Trend Decomposition for Enhanced Time Series Forecasting Authors:Sanjeev Panta, Xu Yuan, Li Chen, Nian-Feng Tzeng View a PDF of the paper titled Revisiting the Seasonal Trend Decomposition for Enhanced Time Series Forecasting, by Sanjeev Panta and 3 other authors View PDF HTML (experimental) Abstract:Time series forecasting presents significant challenges in real-world applications across various domains. Building upon the decomposition of the time series, we enhance the architecture of machine learning models for better multivariate time series forecasting. To achieve this, we focus on the trend and seasonal components individually and investigate solutions to predict them with less errors. Recognizing that reversible instance normalization is effective only for the trend component, we take a different approach with the seasonal component by directly applying backbone models without any normalization or scaling procedures. Through these strategies, we successfully reduce error values of the existing state-of-the-art models and finally introduce dual-MLP models as more computationally efficient solutions. Furthermore, our approach consistently yields positive results with around 10% MSE average reduction across four state-of-the-art baselines on the benchmark datasets. We also evaluate our approach on a hydrological dataset extracted from the United States Geol...
