An application of the ensemble Kalman filter in epidemiological modelling

Rajnesh Lal, Weidong Huang, Zhenquan Li

Research output: Contribution to journalArticlepeer-review

Abstract

Since the novel coronavirus (COVID-19) outbreak in China, and due to the open accessibility of COVID-19 data, several researchers and modellers revisited the classical epidemiological models to evaluate their practical applicability. While mathematical compartmental models can predict various contagious viruses’ dynamics, their efficiency depends on the model parameters. Recently, several parameter estimation methods have been proposed for different models. In this study, we evaluated the Ensemble Kalman filter’s performance (EnKF) in the estimation of time-varying model parameters with synthetic data and the real COVID-19 data of Hubei province, China. Contrary to the previous works, in the current study, the effect of damping factors on an augmented EnKF is studied. An augmented EnKF algorithm is provided, and we present how the filter performs in estimating models using uncertain observational (reported) data. Results obtained confirm that the augumented-EnKF approach can provide reliable model parameter estimates. Additionally, there was a good fit of profiles between model simulation and the reported COVID-19 data confirming the possibility of using the augmented-EnKF approach for reliable model parameter estimation.
Original languageEnglish
Article number0256227
Pages (from-to)1-25
Number of pages25
JournalPLoS One
Volume16
Issue number8
DOIs
Publication statusPublished - 19 Aug 2021

Fingerprint

Dive into the research topics of 'An application of the ensemble Kalman filter in epidemiological modelling'. Together they form a unique fingerprint.

Cite this