Abstract:In real-world scenarios like traffic and energy, massive time-series data with missing values and noises are widely observed, even sampled irregularly. While many imputation methods have been proposed, most of them work with a local horizon, which means models are trained by splitting the long sequence into batches of fit-sized patches. This local horizon can make models ignore global trends or periodic patterns. More importantly, almost all methods assume the observations are sampled at regular time stamps, and fail to handle complex irregular sampled time series arising from different applications. Thirdly, most existing methods are learned in an offline manner. Thus, it is not suitable for many applications with fast-arriving streaming data. To overcome these limitations, we propose BayOTIDE: Bayesian Online Multivariate Time series Imputation with functional decomposition. We treat the multivariate time series as the weighted combination of groups of low-rank temporal factors with different patterns. We apply a group of Gaussian Processes (GPs) with different kernels as functional priors to fit the factors. For computational efficiency, we further convert the GPs into a state-space prior by constructing an equivalent stochastic differential equation (SDE), and developing a scalable algorithm for online inference. The proposed method can not only handle imputation over arbitrary time stamps, but also offer uncertainty quantification and interpretability for the downstream application. We evaluate our method on both synthetic and real-world datasets.We release the code at {<a class="link-external link-https" href="https://github.com/xuangu-fang/BayOTIDE" rel="external noopener nofollow">this https URL</a>}

An Adaptive Data-Driven Imputation Model for Incomplete Event Series.

Iterative Time Series Imputation by Maintaining Dependency Consistency

An Empirical Study: Extensive Deep Temporal Point Process

Time-series Imputation of Temporally-occluded Multiagent Trajectories

Learning Conditional Generative Models for Temporal Point Processes

Learning Latent Process from High-Dimensional Event Sequences Via Efficient Sampling

How Do Your Neighbors Disclose Your Information: Social-Aware Time Series Imputation

Imitative Non-Autoregressive Modeling for Trajectory Forecasting and Imputation

Multiple-level Point Embedding for Solving Human Trajectory Imputation with Prediction

Networked Time-series Prediction with Incomplete Data via Generative Adversarial Network

BayOTIDE: Bayesian Online Multivariate Time series Imputation with functional decomposition

A time series continuous missing values imputation method based on generative adversarial networks

Recover Missing Sensor Data with Iterative Imputing Network

Multistate time series imputation using generative adversarial network with applications to traffic data

Missing Data Imputation for Industrial Time Series with Adaptive Median Iteration Based on Generative Adversarial Networks

Learning Time Series Associated Event Sequences with Recurrent Point Process Networks

Time Perception Machine: Temporal Point Processes for the When, Where and What of Activity Prediction

Spatio-Temporal View GAIN for Data Imputation and Dynamic Soft Sensor

Missingness-Pattern-Adaptive Learning With Incomplete Data

Uncovering the Missing Pattern: Unified Framework Towards Trajectory Imputation and Prediction

Event prediction and causality inference despite incomplete information