DPWMixer: Dual-Path Wavelet Mixer for Long-Term Time Series Forecasting

TL;DR

DPWMixer introduces a dual-path wavelet-based architecture for long-term time series forecasting, effectively capturing both macro trends and micro dynamics while avoiding information loss common in traditional multi-scale methods.

Contribution

It proposes a novel lossless wavelet pyramid and a dual-path architecture combining linear trend modeling with flexible local dynamics, advancing state-of-the-art performance.

Findings

Consistently outperforms existing methods on eight benchmarks.

Effectively disentangles trends and fluctuations without information loss.

Achieves better long-term forecasting accuracy.

Abstract

Long-term time series forecasting (LTSF) is a critical task in computational intelligence. While Transformer-based models effectively capture long-range dependencies, they often suffer from quadratic complexity and overfitting due to data sparsity. Conversely, efficient linear models struggle to depict complex non-linear local dynamics. Furthermore, existing multi-scale frameworks typically rely on average pooling, which acts as a non-ideal low-pass filter, leading to spectral aliasing and the irreversible loss of high-frequency transients. In response, this paper proposes DPWMixer, a computationally efficient Dual-Path architecture. The framework is built upon a Lossless Haar Wavelet Pyramid that replaces traditional pooling, utilizing orthogonal decomposition to explicitly disentangle trends and local fluctuations without information loss. To process these components, we design a Dual-Path Trend Mixer that integrates a global linear mapping for macro-trend anchoring and a flexible patch-based MLP-Mixer for micro-dynamic evolution. Finally, An adaptive multi-scale fusion module then integrates predictions from diverse scales, weighted by channel stationarity to optimize synthesis. Extensive experiments on eight public benchmarks demonstrate that our method achieves a consistent improvement over state-of-the-art baselines. The code is available at https://github.com/hit636/DPWMixer.