通过频谱分解增强扩散模型的时间序列预测能力
A Decomposable Forward Process in Diffusion Models for Time-Series Forecasting
- 将信号分解为频谱成分，按能量分阶段加噪
- 在扩散过程中保持主频段信噪比，提升长期模式可恢复性
- 兼容现有扩散模型，计算开销极低，适合时序预测任务
我们提出一种面向时间序列预测的模型无关前向扩散过程，该过程将信号分解为频谱成分，相比标准扩散方法更有效地保留季节性等结构化时间模式。不同于以往修改网络架构或直接在频域扩散的工作，本文仅改变扩散过程本身，兼容现有扩散骨干网络（如 DiffWave、TimeGrad、CSDI）。通过按成分能量分阶段注入噪声，维持主频段在整个扩散轨迹中的高信噪比，从而延长信号结构的保持时间，提升长期模式的可恢复性。在多个标准预测基准上，应用傅里叶或小波变换等频谱分解策略均显著优于基线扩散模型，且计算开销可忽略不计。
原文摘要 · Abstract (English)
We introduce a model-agnostic forward diffusion process for time-series forecasting that decomposes signals into spectral components, preserving structured temporal patterns such as seasonality more effectively than standard diffusion. Unlike prior work that modifies the network architecture or diffuses directly in the frequency domain, our proposed method alters only the diffusion process itself, making it compatible with existing diffusion backbones (e.g., DiffWave, TimeGrad, CSDI). By staging noise injection according to component energy, it maintains high signal-to-noise ratios for dominant frequencies throughout the diffusion trajectory, thereby improving the recoverability of long-term patterns. This strategy enables the model to maintain the signal structure for a longer period in the forward process, leading to improved forecast quality. Across standard forecasting benchmarks, we show that applying spectral decomposition strategies, such as the Fourier or Wavelet transform, consistently improves upon diffusion models using the baseline forward process, with negligible computational overhead. The code for this paper is available at https://anonymous.4open.science/r/D-FDP-4A29.
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。