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

By Y. Tashiro, Jiaming Song, Yang Song, Stefano Ermon. Neural Information Processing Systems, 2021

1,078 citations recorded by the source corpus.

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.

Subfields: Artificial Intelligence, Experimental and Cognitive Psychology, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/8982bb695dcebdacbfd079c62cd7acca8a8b48dc

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Experimental and Cognitive Psychology

In Statistics and Probability

Citations in the corpus

A Survey on Generative Diffusion Models (2022, methodology): “CSDI proposed a self-supervised training framework based on bidirectional CNN modules, achieving substantial improvement in continuous generation of healthcare and environmental data [17].”
Diffusion Models: A Comprehensive Survey of Methods and Applications (2022, methodology): “The procedure of time series imputation with CSDI [230].”
Flexible Diffusion Modeling of Long Videos (2022, methodology): “We will describe the conditional extension [31], in which the modeled x is conditioned on observations y.”
Diffusion-based Time Series Imputation and Forecasting with Structured State Space Models (2022, methodology): “[46], we base our parameterization of θ(xt, t, c) on the DiffWave architecture [26].”
Diffusion models as plug-and-play priors (2022, methodology): “DDPMs have previously been used for conditional generation and image segmentation [34, 39, 1].”
TabDDPM: Modelling Tabular Data with Diffusion Models (2022, background): “…such as NLP (Austin et al., 2021; Li et al., 2022), waveform signal processing (Kong et al., 2020; Chen et al., 2020b), molecular graphs (Jing et al., 2022; Hooge-boom et al., 2022), time series (Tashiro et al., 2021), testifying the universality of diffusion models across a wide range of problems.”
eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers (2022, background): “They have driven successful applications such as text-to-image generation [57, 59, 63], natural language generation [41], time series prediction [76], audio synthesis [39], 3D shape generation [49, 87, 91], molecular conformation generation [84], protein structure generation [83], machine learning s”
DiffuSeq: Sequence to Sequence Text Generation with Diffusion Models (2022, background): “There are other conditional diffusion models on the time series prediction like CSDI (Tashiro et al., 2021) or audio generation like WaveGrad (Chen et al.”
Membership Inference Attacks against Generative Models with Probabilistic Fluctuation (positive): “…by abundant data resources and em-powered by deep neural networks, generative models have achieved remarkable success in various domains such as computer vision [Rombach et al. , 2022], natural language processing [Li et al. , 2021], and spatial-temporal data modeling [Tashiro et al. , 2021].”