An Intelligent Clustering Approach For Analyzing A Multivariate Time Series Dataset, Case Study COVID-19 Outbreak in Indonesia

The coronavirus disease 2019 (COVID-19) became a pandemic across the world in 2020 - 2022. Indonesia is one of the countries affected by the COVID-19 outbreak. As an archipelago country, the spreading of COVID-19 in one and other provinces varies. This study is conducted to cluster the provinces in Indonesia according to the daily new cases, deaths, and recoveries of COVID-19 patients. It implements Intelligent Kernel K-Means and Intelligent K-Means to cluster multivariate time series data without determining the number of clusters. Intelligent Kernel K-Means finds two clusters and obtains the best average Silhouette score of 0.81. It also reveals that the severity of COVID-19 in some areas correlated to the variables population density, population, general hospitals, inpatient health centers, not inpatient health centers, urban areas with water pollution, urban areas with land pollution, urban areas with air pollution, and urban areas without pollution.