科研成果详情

Aimed to provide insight into the global trend of COVID-19 during the Delta variant stage and offer recommendations for effective epidemic prevention policies. In this study, we propose a novel data-driven causal inference model combining fuzzy c-means (FCM) clustering and difference-in-difference (DID) for multiple time series data. Based on the FCM clustering, a set of panel data with parallel trends can be obtained for further DID analysis. By comparing the change in the outcome variable between the treatment and control groups within each cluster, before and after the intervention, the causal effect of each cluster on the outcome variable is estimated. In government response analysis to COVID-19, the daily updated data of 196 countries all over the world during the transition phase from the outbreak of the SARS-CoV-2 Delta variant to the discovery of Omicron was collected. Application of our method shows that adopting restrictions on internal movement policies had a substantial impact on preventing the spread of the pandemic. The FCM-DID approach we proposed is useful for identifying and comparing the effects of policies on the outcome of different groups, thus can well evaluate the effectiveness of implemented policies and provide valuable guidance for decision-making in future public health crises.