arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.17706cs.LG
收藏

并行复数扩散用于可扩展时间序列生成

Parallel Complex Diffusion for Scalable Time Series Generation

Institute of Cyber-Systems and Control(网络系统与控制研究所);Zhejiang University(浙江大学);Griffith University(格里菲斯大学);School of Mathematics(数学学院);Southeast University(东南大学);Squirrel Ai Learning

Rongyao Cai, Yuxi Wan, Kexin Zhang, Ming Jin, Zhiqiang Ge, Qingsong Wen, Yong Liu

AI总结:

提出PaCoDi(并行复数扩散)框架,通过离散傅里叶变换将时间序列分解到谱域,利用并行实值估计器替代复数估计器,解决时间序列生成中的纠缠问题,理论证明谱高斯噪声的正交性,并引入平均场理论近似处理边缘耦合,在无条件和条件生成任务中优于5个基线。

摘要:

扩散模型通过去噪间接学习数据分布,使得生成建模的难度与数据的依赖结构密切相关。对于时间序列,强时间依赖性迫使噪声/分数估计器恢复高度纠缠的跨时间关系,导致纠缠问题。我们通过改变扩散空间的拓扑结构来减轻这一负担:离散傅里叶变换(DFT)将时间依赖分解为谱模式,对角化二阶依赖结构,使数据流形与各向同性高斯噪声和均匀扩散动力学更好地对齐。然而,现有的频率感知扩散方法主要使用DFT设计时间DDPM/SDE框架下的估计器模块,而频率原生扩散路径面临复数动力学带来的数学障碍。我们提出PaCoDi(并行复数扩散),一种频率原生扩散框架,在谱域构建扩散路径,同时用实部和虚部的并行实值估计器替代复数估计器。理论上,我们证明了谱高斯噪声的统计正交性,建立了正交前向转移和条件反向分解,并通过谱维纳过程将离散PaCoDi扩展到连续时间谱SDE。我们进一步引入带有交互校正分支的平均场理论近似来处理边缘耦合,并利用厄米对称性减少50%的注意力FLOPs而无信息损失。在无条件和条件时间序列生成上的大量实验表明,在5个基准测试中,生成质量和计算效率分别优于5个最先进基线。代码可在https://github.com/RongyaoCai/PaCoDi获取。

英文摘要:

Diffusion models learn data distributions indirectly through denoising, making the difficulty of generative modeling closely tied to the dependency structure of data. For time series, strong temporal dependence forces the noise / score estimator to recover highly entangled cross-time relationships, leading to the curse of entanglement. We mitigate this burden by changing the topology of the diffusion space: the Discrete Fourier Transform (DFT) decomposes temporal dependencies into spectral modes, diagonalizing second-order dependency structure and better aligning the data manifold with isotropic Gaussian noise and homogeneous diffusion dynamics. However, existing frequency-aware diffusion methods mainly use the DFT to design estimator blocks under temporal DDPM/SDE frameworks, while frequency-native diffusion paths face a mathematical barrier from complex-valued dynamics. We propose PaCoDi (Parallel Complex Diffusion), a frequency-native diffusion framework that constructs the diffusion path in the spectral domain while replacing the complex-valued estimator with parallel real-valued estimators for real and imaginary components. Theoretically, we prove the statistical orthogonality of spectral Gaussian noise, establish quadrature forward transitions and conditional reverse factorization, and extend discrete PaCoDi to continuous-time spectral SDEs through a Spectral Wiener Process. We further introduce a Mean Field Theory approximation with an Interactive Correction Branch to handle marginal coupling, and exploit Hermitian symmetry to reduce 50% attention FLOPs without information loss. Extensive experiments on unconditional and conditional time series generation demonstrate superior generative quality and computational efficiency against 5 SOTA baselines in 5 benchmarks, respectively. Code is available at https://github.com/RongyaoCai/PaCoDi.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑