A0926
Title: Predicting COVID-19 hospitalization using a mixture of Bayesian predictive syntheses
Authors: Genya Kobayashi - School of Commerce, Meiji University (Japan) [presenting]
Shonosuke Sugasawa - Keio University (Japan)
Yuki Kawakubo - Chiba University (Japan)
Taeryon Choi - Korea University (Korea, South)
Dongu Han - Korea University (Korea, South)
Abstract: A novel methodology is proposed, called the mixture of Bayesian predictive syntheses (MBPS), for multiple time series count data for the challenging task of predicting the numbers of COVID-19 inpatients and isolated cases in Japan and Korea at the subnational level. MBPS combines a set of predictive models and partitions the multiple time series into clusters based on their contribution to predicting the outcome. In this way, MBPS leverages the shared information within each cluster and is suitable for predicting COVID-19 inpatients since the data exhibit similar dynamics over multiple areas. Also, MBPS avoids using a multivariate count model, which is generally cumbersome to develop and implement. The Japanese and Korean data analyses demonstrate that the proposed MBPS methodology has improved predictive accuracy and uncertainty quantification.
