Time-series imputation using low-rank matrix completion

TL;DR

This paper explores using low-rank matrix completion on block-Hankel matrices for time-series imputation, demonstrating competitive performance especially in capturing sharp peaks compared to existing methods.

Contribution

It introduces a novel application of low-rank matrix completion to time-series imputation using block-Hankel matrices, showing its effectiveness in reproducing sharp data features.

Findings

Hankel Imputation (HI) performs competitively in interpolation tasks.

HI is particularly effective at reproducing sharp peaks.

The method offers a balance of computational effort and imputation quality.

Abstract

We investigate the use of matrix completion methods for time-series imputation. Specifically we consider low-rank completion of the block-Hankel matrix representation of a time-series. Simulation experiments are used to compare the method with five recognised imputation techniques with varying levels of computational effort. The Hankel Imputation (HI) method is seen to perform competitively at interpolating missing time-series data, and shows particular potential for reproducing sharp peaks in the data.