Diffusion models for time-series applications: a survey

2024, Volume 25, Issue 1

Abstract

Keywords

Related Research

Frontiers of Information Technology & Electronic Engineering >> 2024, Volume 25, Issue 1 doi: 10.1631/FITEE.2300310

Diffusion models for time-series applications: a survey

1. Discipline of Business Analytics, The University of Sydney Business School, Camperdown, NSW 2006, Australia; 2. Postdoctoral Programme of Zhongtai Securities Co., Ltd., Jinan 250000, China;

Received: 2023-04-29 Accepted: 2024-02-19 Available online: 2024-02-19

HTML0 PDF 51 Collect 0

Next Previous

Abstract

, a family of generative models based on deep learning, have become increasingly prominent in cutting-edge machine learning research. With distinguished performance in generating samples that resemble the observed data, are widely used in image, video, and text synthesis nowadays. In recent years, the concept of diffusion has been extended to time-series applications, and many powerful models have been developed. Considering the deficiency of a methodical summary and discourse on these models, we provide this survey as an elementary resource for new researchers in this area and to provide inspiration to motivate future research. For better understanding, we include an introduction about the basics of . Except for this, we primarily focus on diffusion-based methods for , imputation, and generation, and present them, separately, in three individual sections. We also compare different methods for the same application and highlight their connections if applicable. Finally, we conclude with the common limitation of diffusion-based methods and highlight potential future research directions.

Keywords

Diffusion models ; Time-series forecasting ; Time-series imputation ; Denoising diffusion probabilistic models ; Score-based generative models ; Stochastic differential equations

Related Research