CMoS: Rethinking Time Series Prediction Through the Lens of Chunk-wise Spatial Correlations

2025/05/25 by Si, Haotian, Pei, Changhua, Li, Jianhui +2 · 2 citations
#FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML)

paper · doi:10.48550/arxiv.2505.19090

Abstract

Recent advances in lightweight time series forecasting models suggest the inherent simplicity of time series forecasting tasks. In this paper, we present CMoS, a super-lightweight time series forecasting model. Instead of learning the embedding of the shapes, CMoS directly models the spatial correlations between different time series chunks. Additionally, we introduce a Correlation Mixing technique that enables the model to capture diverse spatial correlations with minimal parameters, and an optional Periodicity Injection technique to ensure faster convergence. Despite utilizing as low as 1% of the lightweight model DLinear's parameters count, experimental results demonstrate that CMoS outperforms existing state-of-the-art models across multiple datasets. Furthermore, the learned weights of CMoS exhibit great interpretability, providing practitioners with valuable insights into temporal structures within specific application scenarios.

Citations

• CycleNet: Enhancing Time Series Forecasting through Modeling Periodic Patterns
• SparseTSF: Modeling Long-term Time Series Forecasting with 1k Parameters
• SOFTS: Efficient Multivariate Time Series Forecasting with Series-Core Fusion
• TFB: Towards Comprehensive and Fair Benchmarking of Time Series Forecasting Methods
• MOMENT: A Family of Open Time-series Foundation Models
• Unified Training of Universal Time Series Forecasting Transformers
• Timer: Generative Pre-trained Transformers Are Large Time Series Models
• iTransformer: Inverted Transformers Are Effective for Time Series Forecasting
• SegRNN: Segment Recurrent Neural Network for Long-Term Time Series Forecasting
• FITS: Modeling Time Series with 10k Parameters
• Long-term Forecasting with TiDE: Time-series Dense Encoder
• A Time Series is Worth 64 Words: Long-term Forecasting with Transformers
• TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysis
• Less Is More: Fast Multivariate Time Series Forecasting with Light Sampling-oriented MLP Structures
• Are Transformers Effective for Time Series Forecasting?
• FEDformer: Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting
• Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting
• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• N-BEATS: Neural basis expansion analysis for interpretable time series forecasting
• Adaptive Mixtures of Local Experts

Cited by

• SVTime: Small Time Series Forecasting Models Informed by "Physics" of Large Vision Model Forecasters
• Bridging Past and Future: Distribution-Aware Alignment for Time Series Forecasting

Related

• LightCTS: A Lightweight Framework for Correlated Time Series Forecasting
• AutoCTS: Automated Correlated Time Series Forecasting -- Extended Version
• AverageTime: Enhance Long-Term Time Series Forecasting with Simple Averaging
• A Comprehensive Survey of Deep Learning for Time Series Forecasting: Architectural Diversity and Open Challenges
• Decomposing the Time Series Forecasting Pipeline: A Modular Approach for Time Series Representation, Information Extraction, and Projection