TimeBridge: Better Diffusion Prior Design with Bridge Models for Time Series Generation

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Park, Jinseong, Lee, Seungyun, Jeong, Woojin, Choi, Yujin, Lee, Jaewook
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866915693078773760
author Park, Jinseong
Lee, Seungyun
Jeong, Woojin
Choi, Yujin
Lee, Jaewook
author_facet Park, Jinseong
Lee, Seungyun
Jeong, Woojin
Choi, Yujin
Lee, Jaewook
contents Time series generation is widely used in real-world applications such as simulation, data augmentation, and hypothesis testing. Recently, diffusion models have emerged as the de facto approach to time series generation, enabling diverse synthesis scenarios. However, the fixed standard-Gaussian diffusion prior may be ill-suited for time series data, which exhibit properties such as temporal order and fixed time points. In this paper, we propose TimeBridge, a framework that flexibly synthesizes time series data by using diffusion bridges to learn paths between a chosen prior and the data distribution. We then explore several prior designs tailored to time series synthesis. Our framework covers (i) data- and time-dependent priors for unconditional generation and (ii) scale-preserving priors for conditional generation. Experiments show that our framework with data-driven priors outperforms standard diffusion models on time series generation.
format Preprint
id arxiv_https___arxiv_org_abs_2408_06672
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle TimeBridge: Better Diffusion Prior Design with Bridge Models for Time Series Generation
Park, Jinseong
Lee, Seungyun
Jeong, Woojin
Choi, Yujin
Lee, Jaewook
Machine Learning
Artificial Intelligence
Time series generation is widely used in real-world applications such as simulation, data augmentation, and hypothesis testing. Recently, diffusion models have emerged as the de facto approach to time series generation, enabling diverse synthesis scenarios. However, the fixed standard-Gaussian diffusion prior may be ill-suited for time series data, which exhibit properties such as temporal order and fixed time points. In this paper, we propose TimeBridge, a framework that flexibly synthesizes time series data by using diffusion bridges to learn paths between a chosen prior and the data distribution. We then explore several prior designs tailored to time series synthesis. Our framework covers (i) data- and time-dependent priors for unconditional generation and (ii) scale-preserving priors for conditional generation. Experiments show that our framework with data-driven priors outperforms standard diffusion models on time series generation.
title TimeBridge: Better Diffusion Prior Design with Bridge Models for Time Series Generation
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2408.06672