xPatch用双流结构+指数分解提升时序预测精度
xPatch: Dual-Stream Time Series Forecasting with Exponential Seasonal-Trend Decomposition
- 双流设计：线性流用MLP，非线性流用CNN，分离处理趋势与季节性
- 引入指数季节-趋势分解，更好捕捉时间序列的动态模式
- 适合需要高精度时序预测的工业或金融场景
近年来，基于Transformer的时序预测模型受到广泛关注。尽管表现优异，但其注意力机制难以充分挖掘时间序列中的时序关系。本文提出eXponential Patch（xPatch），一种新型双流架构，采用指数分解方法。受经典指数平滑启发，xPatch引入季节-趋势指数分解模块。同时设计双流结构，包含基于MLP的线性流和基于CNN的非线性流。该模型探索了在非Transformer框架下使用分块（patching）与通道独立性的优势。最后，提出鲁棒的反正切损失函数与Sigmoid学习率调整策略，有效防止过拟合并提升预测性能。代码已开源：https://github.com/stitsyuk/xPatch。
原文摘要 · Abstract (English)
In recent years, the application of transformer-based models in time-series forecasting has received significant attention. While often demonstrating promising results, the transformer architecture encounters challenges in fully exploiting the temporal relations within time series data due to its attention mechanism. In this work, we design eXponential Patch (xPatch for short), a novel dual-stream architecture that utilizes exponential decomposition. Inspired by the classical exponential smoothing approaches, xPatch introduces the innovative seasonal-trend exponential decomposition module. Additionally, we propose a dual-flow architecture that consists of an MLP-based linear stream and a CNN-based non-linear stream. This model investigates the benefits of employing patching and channel-independence techniques within a non-transformer model. Finally, we develop a robust arctangent loss function and a sigmoid learning rate adjustment scheme, which prevent overfitting and boost forecasting performance. The code is available at the following repository: https://github.com/stitsyuk/xPatch.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。