NPMixer: Hierarchical Neighboring Patch Mixing for Time Series Forecasting
For practitioners of time series forecasting, NPMixer offers a new method that consistently beats current SOTA models on multiple benchmarks, though the gains are incremental.
NPMixer introduces a hierarchical architecture with a learnable wavelet transform and neighboring patch mixing to capture both local temporal dynamics and global dependencies in multivariate time series forecasting. It achieves state-of-the-art performance, outperforming existing models in 71.4% of evaluated setups across seven benchmarks.
Multivariate time series forecasting remains a challenge due to the complexity of local temporal dynamics and global dependencies across multiple variables. In this paper, we propose \textbf{N}eighboring \textbf{P}atching \textbf{Mixer} (\textbf{NPMixer}), a hierarchical architecture featuring a Learnable Stationary Wavelet Transform that adaptively learns filter coefficients to decompose signals into trend and detail components in a data-dependent manner. Our framework introduces a Neighboring Mixer Block that captures local temporal dynamics through a series of hierarchical MLP layers operating on non-overlapping patches. Specifically, the mixer block utilizes MLPs to learn temporal patterns within and across these patches, expanding the receptive field to capture multi-scale dependencies. A Channel-Mixing Encoder is applied to high-frequency components to learn channel correlations while preserving the stability of the underlying global trend. Extensive experiments on seven benchmark datasets demonstrate that NPMixer consistently outperforms state-of-the-art models, achieving better performance in 20 out of 28 ($71.4\%$) evaluated experimental setups for MSE.