Scalable Hot-spot Detection for the Korean COVID-19 Trajectory Data Using Kernel Density Estimation

COVID-19 has been soaring as a destructive pandemic. During this epidemic disaster, it is a fundamental process to effectively use quarantine-resources that finding Hot-spot using contact-tracing data. Since the first case of COVID-19 identified in South Korea, the trajectory information and its details of confirmed COVID-19 cases have been recorded from publicly available resources. In this study, we apply weighted Kernel density estimation(KDE), the non-parametric density estimator, to discovering spatial and critical locations that can be a potential infection occurring. We use Korean trajectory-tracing data and the number of contacted individuals for each patient as the weight of KDE. As a result, we show the visualization of our model for the trajectory data. We examine the effect of contacted numbers to adjust the fixed weight related to the number of contact individuals.