Abstract
Many multivariate time series observed in practice are second order nonstationary, i.e. their covariance properties vary over time. In addition, missing observations in such data are encountered in many applications of interest, due to recording failures or sensor dropout, hindering successful analysis. This article introduces a novel method for data imputation in multivariate nonstationary time series, based on the so-called locally stationary wavelet modelling paradigm. Our methodology is shown to perform well across a range of simulation scenarios, with a variety of missingness structures, as well as being competitive in the stationary time series setting. We also demonstrate our technique on data arising in a health monitoring application.
| Original language | English |
|---|---|
| Article number | 18 |
| Journal | Statistics and Computing |
| Volume | 31 |
| Issue number | 2 |
| DOIs | |
| Publication status | Published - 17 Feb 2021 |
Funding
Wilson gratefully acknowledges financial support from EPSRC and Shell via the STOR-i Centre for Doctoral Training (EP/L015692/1). Eckley also gratefully acknowledges the financial support of EPSRC Grant EP/N031938/1. An R package implementing the imputation method in this article is available from the corresponding author on request.
Keywords
- imputation; local stationarity; missing data; multivariate time series