Hybrid clustering and classification methods to find out the pattern of the spread of covid-19 in East Java province

Abstract

Abstract Covid-19 entered Indonesia in March 2020, which had an extraordinary impact on all aspects of life. The real impact is the economic and social sectors. Various efforts and policies from the Indonesian government were made to suppress the increase in the rate of Covid-19 cases. The spread of Covid-19 cases in Indonesia has been evenly distributed in various provinces, one of which is East Java. East Java has 38 districts/cities, where the case of each district is different, there are green, yellow, red, and black zones. To find out the pattern of spread of each district/city zone, clustering is used. The method used in this study was to use the hybrid method of clustering with K-means and classification with its algorithm Decision Tree. The purpose of this study to prepare mitigation measures to inhibit the rate of spread of Covid-19 in East Java Province. The variables used include the number of positive, the number of dead, the number of recovered, the number of suspects, and the number of probable. With this grouped data, it is expected to help make the right decision in reducing the spread and minimizing the number of positive patients. Based on the results of the study, there will be cluster results, namely 7 very high-risk areas (C0), 1 high risk district (C1), 29 moderate risk districts (C2), and 1 low-risk district (C3). It also has a validity index level of 0.356 as measured by the Davies-Bouldin Index.