CSDI: Conditional Score-based Diffusion Models for Probabilistic Time Series Imputation

Abstract

The imputation of missing values in time series has many applications in healthcare and finance. While autoregressive models are natural candidates for time series imputation, score-based diffusion models have recently outperformed existing counterparts including autoregressive models in many tasks such as image generation and audio synthesis, and would be promising for time series imputation. In this paper, we propose Conditional Score-based Diffusion models for Imputation (CSDI), a novel time series imputation method that utilizes score-based diffusion models conditioned on observed data. Unlike existing score-based approaches, the conditional diffusion model is explicitly trained for imputation and can exploit correlations between observed values. On healthcare and environmental data, CSDI improves by 40-65% over existing probabilistic imputation methods on popular performance metrics. In addition, deterministic imputation by CSDI reduces the error by 5-20% compared to the state-of-the-art deterministic imputation methods. Furthermore, CSDI can also be applied to time series interpolation and probabilistic forecasting, and is competitive with existing baselines. The code is available at https://github.com/ermongroup/CSDI.

arXiv commentNeurIPS 2021

References in corpus (8)

• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• A Gradient Descent Algorithm on the Grassman Manifold for Matrix Completion
• Improved Denoising Diffusion Probabilistic Models
• Autoregressive Denoising Diffusion Models for Multivariate Probabilistic Time Series Forecasting
• High-Dimensional Multivariate Forecasting with Low-Rank Gaussian Copula Processes
• Solving Linear Inverse Problems Using the Prior Implicit in a Denoiser
• Symbolic Music Generation with Diffusion Models
• Temporal Latent Auto-Encoder: A Method for Probabilistic Multivariate Time Series Forecasting

Cited by in corpus (8)

• Spatio-temporal Diffusion Point Processes
• MTSCI: A Conditional Diffusion Model for Multivariate Time Series Consistent Imputation
• Self-Supervised Learning of Time Series Representation via Diffusion Process and Imputation-Interpolation-Forecasting Mask
• Graph Neural Processes for Spatio-Temporal Extrapolation
• How Much is Enough? A Study on Diffusion Times in Score-based Generative Models
• Deep Generative Imputation Model for Missing Not At Random Data
• Causality-Aware Spatiotemporal Graph Neural Networks for Spatiotemporal Time Series Imputation
• Self-Supervision Improves Diffusion Models for Tabular Data Imputation