Abstract:The stable periodic patterns present in time series data serve as the foundation for conducting long-horizon forecasts. In this paper, we pioneer the exploration of explicitly modeling this periodicity to enhance the performance of models in long-term time series forecasting (LTSF) tasks. Specifically, we introduce the Residual Cycle Forecasting (RCF) technique, which utilizes learnable recurrent cycles to model the inherent periodic patterns within sequences, and then performs predictions on the residual components of the modeled cycles. Combining RCF with a Linear layer or a shallow MLP forms the simple yet powerful method proposed in this paper, called CycleNet. CycleNet achieves state-of-the-art prediction accuracy in multiple domains including electricity, weather, and energy, while offering significant efficiency advantages by reducing over 90% of the required parameter quantity. Furthermore, as a novel plug-and-play technique, the RCF can also significantly improve the prediction accuracy of existing models, including PatchTST and iTransformer. The source code is available at: <a class="link-external link-https" href="https://github.com/ACAT-SCUT/CycleNet" rel="external noopener nofollow">this https URL</a>.

What problem does this paper attempt to address?

### The Problem the Paper Attempts to Solve This paper attempts to improve the performance of Long-term Time Series Forecasting (LTSF) tasks by explicitly modeling periodic patterns in time series data. Specifically, the authors propose a technique called **Residual Cycle Forecasting (RCF)**, which utilizes learnable cyclic periods to model intrinsic periodic patterns in time series data and then predicts the residual part of these periodic patterns. By combining the RCF technique with linear layers or shallow multilayer perceptrons (MLP), a simple yet powerful method called **CycleNet** is formed. ### Key Contributions 1. **Identifying and Explicitly Modeling Periodic Patterns**: - The authors found that stable periodic patterns in time series data are crucial for long-term forecasting tasks. They propose to improve model prediction performance by explicitly modeling these periodic patterns. 2. **Introducing the RCF Technique**: - The RCF technique uses learnable cyclic periods to explicitly model periodic patterns in time series data and then predicts the residual part of these periodic patterns. The RCF technique significantly enhances the performance of basic (or existing) models. 3. **Proposing the CycleNet Model**: - By combining the RCF technique with simple linear layers or shallow MLP, a simple, efficient, and powerful time series forecasting method called CycleNet is formed. CycleNet achieves state-of-the-art performance in multiple domains and has significant efficiency advantages. ### Solution - **RCF Technique**: - **Periodic Pattern Modeling**: First, model the periodic patterns of the sequence in independent channels through learnable cyclic periods. - **Residual Prediction**: Then, predict the residual part of the modeled periodic patterns. - **CycleNet Model**: - **Architecture**: CycleNet consists of the RCF technique and a simple linear layer or shallow MLP. - **Performance**: CycleNet achieves state-of-the-art performance in long-term time series forecasting tasks across multiple domains, with a parameter reduction of over 90%, providing significant efficiency advantages. ### Experimental Results - **Main Results**: - CycleNet achieves state-of-the-art performance on multiple benchmark datasets, particularly in fields such as electricity, weather, and energy. - CycleNet/MLP ranks first overall, and CycleNet/Linear ranks second. - **Efficiency Analysis**: - CycleNet significantly outperforms other mainstream models in terms of parameter count and computational load, especially when compared to models like iTransformer, with a reduction of over tenfold in both parameters and computational load. - **Ablation Study**: - The RCF technique significantly improves the prediction accuracy of basic models (such as linear layers and MLP) on various benchmark datasets, with an improvement margin of approximately 10% to 20%. - The RCF technique can also significantly enhance the prediction accuracy of existing complex models (such as PatchTST and iTransformer). Even for DLinear, which already uses the classic MOV based on STD technique, RCF can provide an improvement of about 20%. ### Conclusion By explicitly modeling periodic patterns in time series data, CycleNet achieves state-of-the-art performance in long-term time series forecasting tasks across multiple domains and has significant efficiency advantages. The RCF technique, as a flexible plug-in module, can significantly enhance the prediction accuracy of existing models.

CycleNet: Enhancing Time Series Forecasting through Modeling Periodic Patterns