TS-Diffusion: Generating Highly Complex Time Series with Diffusion Models

TL;DR

TS-Diffusion is a novel generative model that effectively synthesizes complex irregular, incomplete, and high-dimensional time series data using a combination of neural ODEs and diffusion techniques.

Contribution

The paper introduces TS-Diffusion, a comprehensive framework that handles irregularities, missing data, and high dimensionality in time series generation with a novel combination of neural ODEs and diffusion models.

Findings

TS-Diffusion outperforms previous models on multiple datasets.

The model effectively captures irregularities and missingness.

It generates high-quality, complex time series data.

Abstract

While current generative models have achieved promising performances in time-series synthesis, they either make strong assumptions on the data format (e.g., regularities) or rely on pre-processing approaches (e.g., interpolations) to simplify the raw data. In this work, we consider a class of time series with three common bad properties, including sampling irregularities, missingness, and large feature-temporal dimensions, and introduce a general model, TS-Diffusion, to process such complex time series. Our model consists of three parts under the framework of point process. The first part is an encoder of the neural ordinary differential equation (ODE) that converts time series into dense representations, with the jump technique to capture sampling irregularities and self-attention mechanism to handle missing values; The second component of TS-Diffusion is a diffusion model that learns from the representation of time series. These time-series representations can have a complex distribution because of their high dimensions; The third part is a decoder of another ODE that generates time series with irregularities and missing values given their representations. We have conducted extensive experiments on multiple time-series datasets, demonstrating that TS-Diffusion achieves excellent results on both conventional and complex time series and significantly outperforms previous baselines.