Diffusion-based Time Series Imputation and Forecasting with Structured State Space Models
This addresses a significant obstacle in real-world data analysis pipelines for time series applications, though it appears incremental as it builds on existing technologies.
The paper tackles the problem of missing value imputation in time series data by proposing SSSD, a model that combines diffusion models and structured state space models, achieving state-of-the-art or better probabilistic imputation and forecasting performance across various datasets and missingness scenarios, including challenging blackout-missing cases.
The imputation of missing values represents a significant obstacle for many real-world data analysis pipelines. Here, we focus on time series data and put forward SSSD, an imputation model that relies on two emerging technologies, (conditional) diffusion models as state-of-the-art generative models and structured state space models as internal model architecture, which are particularly suited to capture long-term dependencies in time series data. We demonstrate that SSSD matches or even exceeds state-of-the-art probabilistic imputation and forecasting performance on a broad range of data sets and different missingness scenarios, including the challenging blackout-missing scenarios, where prior approaches failed to provide meaningful results.