Generic COVID-19 epidemic forecast for Estonia by Gaidai multivariate reliability method
Oleg Gaidai, Jinlu Sheng, Yu Cao, Yan Zhu, Stas Loginov
Abstract
This study aims at accurate spatiotemporal assessment of future epidemic outbreak risks, occurring within any given return period, and within any relevant administrative region of Estonia. Authors have recently developed novel statistical spatiotemporal methodology, that can be applied directly to multivariate raw clinical datasets. Novel Gaidai multivariate reliability methodology, described in this study, being particularly appropriate for multi-regional environmental, biological and public health systems. Advocated methodology has been sufficiently validated in the recent studies, yielding reliable long-term risk forecasts of the future epidemic outbreaks. COVID-19 (SARS-COV-2) daily recorded patient numbers were selected in all impacted Estonia national regions. This study results conclude that recommended spatiotemporal approach may effectively utilize even limited raw clinical datasets, the latter can be useful in a wide range of bioinformatics and public health applications. Confidence intervals have been estimated for predicted epidemiological levels.